Fire Safety Design for Tall Buildings 9780367444525, 9781003009818


353 18 10MB

English Pages [251] Year 2021

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Dedication
Table of Contents
Preface
Acknowledgments
Author
1 Introduction
1.1 Aims and Scope
1.2 Main Fire Safety Design Issues for Tall Buildings
1.3 Structure of the Book
Notes
References
2 Regulatory Requirements and Basic Fire Safety Design Principles
2.1 Introduction
2.2 Fire Incidents and Fire Tests of Tall Buildings Worldwide
2.2.1 Grenfell Tower
2.2.1.1 The New Cladding System
2.2.1.2 Compartment and Evacuation Route for Grenfell Tower
2.2.1.3 Collapse Potential for Grenfell Tower
2.2.1.4 Major Findings from Interim Report of British Research Establishment (2017)
2.2.2 Twin Tower
2.2.3 World Trade Center 7
2.2.4 Other Fire Incidents of Tall Buildings
2.2.4.1 First Interstate Bank Building in Los Angles
2.2.4.2 Plasco Shopping Center, Iran
2.2.4.3 Faculty of Architecture Building, Delft University
2.2.4.4 Windsor Tower, Spain
2.2.5 Cardington Fire Test
2.2.5.1 Introduction of the Test
2.2.5.2 Failure Modes for Buildings in Fire
2.2.6 Discussion
2.3 Current Design Guidance and Regulations to Fire Safety in High-Rise Buildings
2.3.1 British Design Guidance and Regulations
2.3.1.1 Building Regulations 2010—Approved Document B
2.3.1.2 The FSO and Housing Act 2004
2.3.1.3 BS 7974:2019
2.3.1.4 BS 9999:2017
2.3.1.5 BS 5950-8:2003
2.3.1.6 BS 476-20:1987
2.3.1.7 Design Guidelines from IStructE and Steel Construction Institute
2.3.2 Eurocode
2.3.3 Guidelines from International Organization for Standardization
2.3.3.1 ISO 24679-1:2019(en)
2.3.3.2 ISO 16730-1 and ISO 16733-1
2.3.3.3 ISO 834-1:1999
2.3.4 US Design Guidance
2.3.4.1 National Fire Protection Association
2.3.4.2 International Code Council— International Fire Code® (IFC®)
2.3.4.3 American Society for Testing and Materials (ASTM)
2.3.4.4 American Society of Civil Engineers
2.3.4.5 Federal Standards and Guidelines
2.3.5 Chinese Design Guidance
2.3.6 New Zealand Code NZS 3404 Part 1:1997
2.3.7 Australian Code AS 4100:1998
2.4 Basic Principles for Fire Safety of Tall Buildings
2.4.1 Main Design Objective
2.4.2 Main Design Tasks
2.4.3 Structural Fire Design
2.4.3.1 Key Design Tasks in Structural Fire Design
2.4.3.2 Design Approach
2.4.3.3 Pros and Cons of the Two Design Methods
2.4.4 Robustness of the Structure in Fire
2.4.5 Fire Modeling
2.4.5.1 Modeling the Atmosphere Temperature Induced by Fire
2.4.5.2 Modeling the Thermal Response of Load-Bearing Building Elements
2.4.5.3 Summary
References
3 Fundamentals of Fire and Fire Safety Design
3.1 Introduction
3.2 Fire Development Process
3.3 Design Fire Temperature
3.1.1 Standard Fire Temperature–time Curve
3.1.2 The Parametric Temperature–Time Curves
3.1.3 Summary
3.4 Design Fire in a Compartment
3.4.1 Characterization of Compartment
3.4.1.1 Characterization of Fire Enclosure
3.4.1.2 Characterization of Openings
3.4.1.3 Duration of Fire to be Adopted in Design
3.4.2 Fuel-Controlled and Ventilation-Controlled Fire
3.4.3 Long-Cool and Short-Hot Fire
3.4.4 Fully Developed Fire
3.4.5 Localized Fire
3.4.5.1 Calculation of Thermal Action of a Localized Fire from Eurocode
3.4.6 Traveling Fire
3.4.7 Fire Scenarios for Tall Buildings
3.5 Fire Severity
3.6 Fire Load
3.6.1 Fire Load Calculation from Eurocode 1
3.6.2 Fire Load Density from Eurocode 1
3.7 Fire Spread
3.8 Routes of Fire Spread
3.8.1 Horizontal Spread of Fire
3.8.2 Vertical Spread of Fire
3.8.2.1 Fire Spread Through Ducts, Shafts, and Penetrations (Internal)
3.8.2.2 Fire Spread Through Façade
3.9 Structural Fire Design
3.9.1 Determine the Compartment Temperature (Design Fire)
3.9.2 Determine the Thermal Response of Structural Members
3.9.3 Heat Transfer
3.9.3.1 Thermodynamics of Heat Transfer
3.9.3.2 Eurocode Formula to Determine Member Temperature
3.9.4 Material Degradation at Elevated Temperatures
3.9.4.1 Degradation of Steel Material in Fire
3.9.4.2 Degradation of Concrete Material in Fire
3.9.5 Design Values of Material Properties Under Fire
3.9.6 Design of Structural Members in Fire
3.9.6.1 Mechanical Design Approaches of Structural Members in Fire
3.9.6.2 The Acceptance Criteria in Designing Structural Members for Tall Buildings
3.10 Fire Resistance
3.10.1 Methods to Determine Fire Resistance
3.10.2 Fire Resistance Rating
3.10.3 Fire Resistance Test for Load- Bearing Structural Members
3.10.4 Fire Resistance Requirements for Elements of a Tall Building
3.11 Fire Protection Method
3.11.1 Active Control System
3.11.2 Passive Control System
3.11.2.1 Intumescent Paints
3.11.2.2 Spray Fire Protection
3.11.2.3 Board Fire Protection
3.11.3 Fire Resistance Test for Protected Members
References
4 Structural Fire Design Principles for Tall Buildings
4.1 Introduction
4.2 Key Tasks for Structural Fire Design
4.2.1 Building Elements to be Considered in Design for Fire
4.2.2 Design of Structural Members in Fire
4.2.3 Design Procedures
4.3 Fire Resistance Rating for Load-Bearing Structural Members
4.4 Design of Concrete Members in Fire
4.4.1 Thermal Response of Concrete in Fire
4.4.2 Spalling
4.4.2.1 Types of Spalling
4.4.2.2 Prevention of Spalling
4.4.3 Simplified Calculation Methods for Concrete Members from EC2 EN 1992-1-2:2004/ A1:2019 (E), 500°C Isotherm Method
4.4.4 Concrete Cover and Protective Layers
4.5 Design of Steel Members in Fire
4.5.1 Thermal Response of Steel in Fire
4.5.2 The Critical Temperature Method (BS5950, 2003 and EN 1993-1-2 2005)
4.5.2.1 Assumptions
4.5.2.2 Load Ratio (Degree of Utilization)
4.5.2.3 Critical Temperature Method for Constrained Members
4.5.2.4 Critical Temperature Method for the Compression and Unconstrained Members
4.5.2.5 Column Buckling Resistance in Fire
4.5.3 Lateral Torsional Buckling of Steel Beams
4.5.4 Beams in Line with Compartment Walls
4.6 Moment Capacity Approach (Section Method)
4.6.1 Method of Calculation
4.6.1.1 Temperature Profile
4.6.1.2 Reduced Strength of Each Element
4.6.1.3 Reduced Flexural Strength Calculation
4.6.2 Case Study for Flexural Capacity of Reinforced Concrete Beams Using Moment Capacity Approach
4.6.3 Flexural Capacity of Steel Beams Using Moment Capacity Approach
4.7 Design of Composite Beams Under Fire
4.7.1 Resistance of Shear Connection in Fire
4.7.2 Effect of Degree of Shear Connection
4.7.3 Edge Beams in Fire
4.7.4 Case Study of Composite Beam Design in Fire
4.8 Design of Composite Slabs in Fire
4.8.1 Membrane Actions in Fire
4.8.2 Strength Design Composite Slabs
4.8.2.1 Calculation Method Based on Plastic Theory
4.8.2.2 Calculation Method Considering Membrane Action
4.8.3 Insulation Criterion of Composite Slabs
4.8.4 Deformation Design of Composite Slabs in Fire
4.9 Design of Post-Tension Slabs in Fire
4.10 Design of Connections Under Fire
4.11 Design of Beam Openings
4.12 Summary of Structural Fire Design Methods
4.12.1 Comparison of Moment Capacity Method and Critical Temperature Method
4.12.2 Comparison of Three Major Methods
References
5 Typical Fire Safety Design Strategy for Tall Buildings
5.1 Introduction
5.2 Fire Safety Design Objectives and Strategies for Tall Buildings
5.2.1 Design Objectives
5.2.2 Design Strategies
5.2.3 Design Process
5.3 Design Strategy for Tall Buildings in Fire
5.3.1 Prescriptive Design
5.3.2 Performance-Based Design
5.3.2.1 Step 1: Set Fire Safety Goals and Objectives
5.3.2.1 Step 2: Determine Performance Criteria
5.3.2.2 Step 3: Analysis of Fire Scenarios
5.3.2.3 Step 4: Protection Strategy
5.3.2.4 Step 5: Determine Whether the Fire Safety Goals are Met
5.3.3 Summary
5.4 Fire Risk Analysis for Tall Buildings
5.4.1 Qualitative Fire Risk Assessment
5.4.2 Quantitative Fire Risk Assessment
5.5 Deterministic and Probabilistic Assessments to Determine the Worst-Case Fire Scenario
5.5.1 Deterministic Approach
5.5.2 Probabilistic Approach
5.6 Compartment Design
5.6.1 Key Components in a Compartment
5.6.1.1 Fire Doors Design
5.6.1.2 Compartment Wall Design
5.6.1.3 Compartment Floor Design
5.6.2 Fire Stop
5.6.3 Cavity Barrier
5.6.4 Fire Damper
5.6.5 Integrity of Compartmentation in Buildings
5.6.5.1 Measures to Accommodate Movements of Compartment Walls Due to Fire
5.6.5.2 Control Movement of Slab
5.7 Evacuation Route Design
5.7.1 Number of Escapes Routes and Exits
5.7.2 Design of Exits
5.7.3 Exit Route
5.7.4 Travel Distance
5.7.5 Staircases and Elevators
5.7.5.1 Protected Staircases and Elevators
5.7.5.2 Fire Lift Lobby
5.7.5.3 External Escape Stairs
5.7.6 Phased/Progressive Evacuation
5.7.7 Refuge
5.7.8 Clear Sign for Evacuation
5.7.9 Computational Models for Evacuation Simulation
5.8 Emergency Vehicle and Firefighter Access
5.8.1 Equipment for Firefighting
5.8.2 Firefighting Lift, Lobby, Shaft, and Stair
5.9 Resisting Fire Spread Through Building Envelope
5.9.1 Resisting Fire Spread Over External Walls
5.9.2 Fire-Resisting Design for Glazing
5.10 Fire Detection, Alarm, and Communication System
5.10.1 Central Fire Alarm System
5.10.2 Smoke Detections
5.10.3 Smoke Control
5.11 Fire and Smoke Suppression System
5.12 Comparison for Fire Protection System for Tall Buildings Across the World
5.13 Case Study of Fire Safety Deign for Burj Khalifa
5.13.1 Evacuation and Refuge
5.13.2 Firefight Access
5.13.3 Staircase and Elevator
5.13.4 Alarm and Warning System
5.13.5 Fire Suppression
5.13.6 Special Water Supply and Pumping System
5.14 Case Study: Structural Fire Design of the Shard
5.14.1 Introduction of the Project
5.14.2 Structural System
5.14.3 Determine the Worst-Case Fire Scenarios
5.14.4 Design for Fire Resistance
5.15 Structural Framing and Structural System
References
6 Fire Analysis and Modeling
6.1 Introduction
6.2 Determining Compartment Fire
6.2.1 Simplified Models from Eurocode
6.2.1.1 Compartment Fires
6.2.1.2 Localized Fires
6.2.2 Advanced Models
6.2.2.1 Zone Models
6.2.2.2 Limitations of Zone Modeling
6.2.2.3 Computational Fluid Dynamics (CFD) Fire Modeling
6.3 Determining Member Temperature
6.3.1 Simplified Temperature Increase Models from Eurocode
6.3.2 Heat Transfer Using Finite Element Method
6.3.2.1 Theoretical Principles
6.3.2.2 Analysis Software and Modeling Example
6.4 Determining Structural Response of Structural Members in Fire
6.4.1 Multi-Physics Fire Analysis (Thermal Mechanical Coupled Analysis)
® 6.4.1.1 Abaqus®
6.4.1.2 ADINA
6.4.2 Sequentially Coupled Thermal-Stress Analysis
6.4.2.1 Sequentially Coupled Thermal-
Stress Analysis Using Abaqus®
6.4.2.2 Sequentially Coupled Thermal-Stress Analysis Using ANSYS
6.4.2.3 Partial Thermal-Mechanical Analyses in OpenSees
6.4.2.4 Codified Thermal-Mechanical Coupled Analysis
6.5 Probabilistic Method for Fire Safety Design
6.5.1 Reliability-Based Structural Fire Design and Analysis
6.5.1.1 The Basic Reliability Design Principles
6.5.1.2 Reliability-Based Design and Analysis Procedure
6.5.1.3 Case Study for Reliability Analysis for Individual Members
6.5.1.4 Case Study for Reliability Analysis for a Whole Building
6.5.2 Fire Fragility Functions
6.5.2.3 Compartment-Level Fragility Function
6.5.2.4 Building-Level Fragility Function
6.5.3 Other Probabilistic Approaches in Fire Safety Design
6.6 Major Fire Analysis Software
6.6.1 Ozone Software
6.6.2 CFAST
6.6.3 FDS
6.6.4 LS-DYNA
6.6.5 OpenSees
References
7 Preventing Fire-Induced Collapse of Tall Buildings
7.1 Introduction
7.2 Design Objective and Functional Requirement for Structural Stability in Fire
7.3 Importance of Collapse Prevention of Tall Buildings in Fire
7.4 Collapse Mechanism of Tall Buildings in Fire
7.4.1 Factors Affecting Thermal Response and Failure Mechanism of Individual Members
7.4.2 Behavior and Failure Mechanism of Steel Beams in Fire
7.4.2.1 Local Buckling of Beams in Connection Area
7.4.2.2 Excessive Deflection
7.4.3 Behavior and Failure Mechanism of Slabs in Fire
7.4.3.1 Membrane Actions of Slabs
7.4.3.2 Effect of Different Fire Scenarios in Composite Slabs
7.4.3.3 Other Research in Composite Slabs in Fire
7.4.4 Behavior and Failure Mechanism of Steel Column in Fire
7.4.4.1 Change of Column Force in Fire
7.4.4.2 Out Plane Bending of Columns
7.4.4.3 Effect of the Slenderness Ratios
7.4.5 Behavior of Connections
7.4.6 Behavior and Failure Mechanism of Concrete Column in Fire
7.5 Whole-Building Behavior of Tall Buildings in Fire
7.5.1 Research of Fu (2016b)
7.5.2 Twin Tower Collapse (WTC1 and WTC2)
7.5.2.1 Structural Framing for WTC1
7.5.2.2 Reason for The Collapse of WTC1
7.5.3 WTC7
7.5.3.1 Structural Framing for WTC7
7.5.3.2 Reason for The Collapse of WTC7
7.5.4 Cardington Test
7.5.4.1 Severity of The Fire
7.5.4.2 Structural Framing
7.5.5 Other Research in Whole Building Behavior
7.6 Overall Building Stability System Design for Fire
7.6.1 Bracing System
7.6.2 Core Wall Design
7.7 Methods for Mitigating Collapse of Buildings in Fire
References
8 New Technologies and Machine Learning in Fire Safety Design
8.1 Introduction
8.2 New Technologies in Fire Safety
8.2.1 PAVA Alarm Systems
8.2.2 IOT in Fire Safety
8.2.2.1 Fire Safety Sensors and BMS
8.2.2.2 Fire Suppression
8.3 Machine Learning in Fire Safety Design
8.3.1 Machine Learning and Its Application in the Construction Industry
8.3.2 Problems Experienced in the Conventional Structural Fire Analysis Approach
8.3.3 Predicting Failure Patterns of Simple Steel-Framed Buildings in Fire
8.3.3.1 Define Failure Pattern
8.3.3.2 Dataset Generation Using the Monte Carlo Simulation and Random Sampling
8.3.3.3 Training and Testing
8.3.3.4 Failure Pattern Prediction
8.3.3.5 Fire Safety Design and Progressive Collapse Potential Check Based on Prediction Results
8.3.4 Predicting and Preventing Fires with Machine Learning
8.3.5 Machine Learning of Fire Hazard Model Simulations for Use in Probabilistic Safety Assessments at Nuclear Power Plants
8.3.6 Learning Algorithms and Programming Language
8.3.6.1 Learning Algorithms
8.3.6.2 Programming Language
References
9 Post-Fire Damage Assessment
9.1 Introduction
9.2 Post-Fire Damage Assessment
9.2.1 Post-Fire Damage Assessment of Concrete Structure
9.2.1.1 Visual Inspection
9.2.1.2 Schmidt Rebound Hammer
9.2.1.3 Petrographic Analysis
9.2.1.4 Spectrophotometer Investigations
9.2.1.5 Reinforcement Sampling
9.2.1.6 Compression Test
9.2.2 Post-Fire Damage Assessment of Structural Steel Members
9.2.2.1 Methods for Post-Fire Damage Assessment
9.2.2.2 Nondestructive Post-Fire Damage Assessment of Structural Steel Members Using the Leeb Harness Method
References
Index
Recommend Papers

Fire Safety Design for Tall Buildings
 9780367444525, 9781003009818

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Fire Safety Design for Tall Buildings

Fire Safety Design for Tall Buildings

Feng Fu

First edition published 2021 by CRC Press 6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742 and by CRC Press 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN © 2021 Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, LLC The right of Feng Fu to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, access www. copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are not available on CCC please contact [email protected] Trademark Notice: Product or corporate names may be trademarks or registered trademarks and are used only for identification and explanation without intent to infringe. ISBN: 978-0-367-44452-5 (hbk) ISBN: 978-1-003-00981-8 (ebk) Typeset in Sabon by codeMantra

To my Daughter Quan

Contents

Preface Acknowledgments Author 1 Introduction

xix xxi xxiii 1

1.1 Aims and scope 1 1.2 Main fire safety design issues for tall buildings 3 1.3 Structure of the book 3 Notes 5 References 5

2 Regulatory requirements and basic fire safety design principles 2.1 2.2

7

Introduction 7 Fire incidents and fire tests of tall buildings worldwide 7 2.2.1 Grenfell Tower 8 2.2.1.1 The new cladding system 9 2.2.1.2 Compartment and evacuation route for Grenfell Tower 9 2.2.1.3 Collapse potential for Grenfell Tower 14 2.2.1.4 Major findings from Interim Report of British Research Establishment (2017) 14 2.2.2 Twin Tower 15 2.2.3 World Trade Center 7 18 2.2.4 Other fire incidents of tall buildings 18 2.2.4.1 First Interstate Bank building in Los Angles 18 2.2.4.2 Plasco shopping center, Iran 19

vii

viii

Contents

2.2.4.3

2.3

2.4

Faculty of Architecture Building, Delft University 19 2.2.4.4 Windsor Tower, Spain 20 2.2.5 Cardington fire test 20 2.2.5.1 Introduction of the test 20 2.2.5.2 Failure modes for buildings in fire 23 2.2.6 Discussion 25 Current design guidance and regulations to fire safety in high-rise buildings 26 2.3.1 British design guidance and regulations 26 2.3.1.1 Building Regulations 2010—Approved Document B 26 2.3.1.2 The FSO and Housing Act 2004 28 2.3.1.3 BS 7974:2019 30 2.3.1.4 BS 9999:2017 31 2.3.1.5 BS 5950-8:2003 31 2.3.1.6 BS 476-20:1987 31 2.3.1.7 Design guidelines from IStructE and Steel Construction Institute 31 2.3.2 Eurocode 32 2.3.3 Guidelines from International Organization for Standardization 32 2.3.3.1 ISO 24679-1:2019(en) 32 2.3.3.2 ISO 16730-1 and ISO 16733-1 32 2.3.3.3 ISO 834-1:1999 33 2.3.4 US design guidance 33 2.3.4.1 National Fire Protection Association 33 2.3.4.2 International Code Council— International Fire Code ® (IFC ®) 33 2.3.4.3 American Society for Testing and Materials (ASTM) 34 2.3.4.4 American Society of Civil Engineers 34 2.3.4.5 Federal Standards and Guidelines 35 2.3.5 Chinese design guidance 35 2.3.6 New Zealand code NZS 3404 Part 1:1997 35 2.3.7 Australian code AS 4100:1998 36 Basic principles for fire safety of tall buildings 36 2.4.1 Main design objective 36 2.4.2 Main design tasks 37 2.4.3 Structural fire design 37

Contents

ix

2.4.3.1

2.4.4 2.4.5

Key design tasks in structural fire design 38 2.4.3.2 Design approach 38 2.4.3.3 Pros and cons of the two design methods 39 Robustness of the structure in fire 39 Fire modeling 39 2.4.5.1 Modeling the atmosphere temperature induced by fire 39 2.4.5.2 Modeling the thermal response of load-bearing building elements 40 2.4.5.3 Summary 40

References 40

3 Fundamentals of fire and fire safety design 3.1 3.2 3.3

3.4

3.5 3.6

3.7 3.8

43

Introduction 43 Fire development process 43 Design fire temperature 44 3.1.1 Standard fire temperature–time curve 45 3.1.2 The parametric temperature–time curves 45 3.1.3 Summary 46 Design fire in a compartment 46 3.4.1 Characterization of compartment 47 3.4.1.1 Characterization of fire enclosure 47 3.4.1.2 Characterization of openings 48 3.4.1.3 Duration of fire to be adopted in design 48 3.4.2 Fuel-controlled and ventilation-controlled fire 49 3.4.3 Long-cool and short-hot fire 49 3.4.4 Fully developed fire 50 3.4.5 Localized fire 50 3.4.5.1 Calculation of thermal action of a localized fire from Eurocode 51 3.4.6 Traveling fire 53 3.4.7 Fire scenarios for tall buildings 53 Fire severity 53 Fire load 55 3.6.1 Fire load calculation from Eurocode 1 55 3.6.2 Fire load density from Eurocode 1 55 Fire spread 56 Routes of fire spread 56

x

Contents

3.8.1 3.8.2

Horizontal spread of fire 57 Vertical spread of fire 57 3.8.2.1 Fire spread through ducts, shafts, and penetrations (internal) 59 3.8.2.2 Fire spread through façade 59 3.9 Structural fire design 59 3.9.1 Determine the compartment temperature (design fire) 60 3.9.2 Determine the thermal response of structural members 61 3.9.3 Heat transfer 61 3.9.3.1 Thermodynamics of heat transfer 61 3.9.3.2 Eurocode formula to determine member temperature 63 3.9.4 Material degradation at elevated temperatures 65 3.9.4.1 Degradation of steel material in fire 65 3.9.4.2 Degradation of concrete material in fire 66 3.9.5 Design values of material properties under fire 66 3.9.6 Design of structural members in fire 67 3.9.6.1 Mechanical design approaches of structural members in fire 67 3.9.6.2 The acceptance criteria in designing structural members for tall buildings 69 3.10 Fire resistance 69 3.10.1 Methods to determine fire resistance 70 3.10.2 Fire resistance rating 70 3.10.3 Fire resistance test for loadbearing structural members 70 3.10.4 Fire resistance requirements for elements of a tall building 71 3.11 Fire protection method 72 3.11.1 Active control system 72 3.11.2 Passive control system 72 3.11.2.1 Intumescent paints 72 3.11.2.2 Spray fire protection 73 3.11.2.3 Board fire protection 74 3.11.3 Fire resistance test for protected members 74 References 74

Contents

4 Structural fire design principles for tall buildings 4.1 4.2

4.3

4.4

4.5

4.6

Introduction 77 Key tasks for structural fire design 77 4.2.1 Building elements to be considered in design for fire 78 4.2.2 Design of structural members in fire 78 4.2.3 Design procedures 78 Fire resistance rating for load-bearing structural members 79 Design of concrete members in fire 79 4.4.1 Thermal response of concrete in fire 82 4.4.2 Spalling 82 4.4.2.1 Types of spalling 82 4.4.2.2 Prevention of spalling 83 4.4.3 Simplified calculation methods for concrete members from EC2 EN 1992-1-2:2004/ A1:2019 (E), 500°C isotherm method 86 4.4.4 Concrete cover and protective layers 88 Design of steel members in fire 88 4.5.1 Thermal response of steel in fire 88 4.5.2 The critical temperature method (BS5950, 2003 and EN 1993-1-2 2005) 88 4.5.2.1 Assumptions 88 4.5.2.2 Load ratio (degree of utilization) 89 4.5.2.3 Critical temperature method for constrained members 89 4.5.2.4 Critical temperature method for the compression and unconstrained members 90 4.5.2.5 Column buckling resistance in fire 92 4.5.3 Lateral torsional buckling of steel beams 92 4.5.4 Beams in line with compartment walls 93 Moment capacity approach (section method) 93 4.6.1 Method of calculation 93 4.6.1.1 Temperature profile 93 4.6.1.2 Reduced strength of each element 94 4.6.1.3 Reduced flexural strength calculation 94

xi

77

xii

Contents

4.6.2

Case study for flexural capacity of reinforced concrete beams using moment capacity approach 95 4.6.3 Flexural capacity of steel beams using moment capacity approach 97 4.7 Design of composite beams under fire 98 4.7.1 Resistance of shear connection in fire 98 4.7.2 Effect of degree of shear connection 100 4.7.3 Edge beams in fire 101 4.7.4 Case study of composite beam design in fire 101 4.8 Design of composite slabs in fire 101 4.8.1 Membrane actions in fire 102 4.8.2 Strength design composite slabs 102 4.8.2.1 Calculation method based on plastic theory 103 4.8.2.2 Calculation method considering membrane action 104 4.8.3 Insulation criterion of composite slabs 106 4.8.4 Deformation design of composite slabs in fire 107 4.9 Design of post-tension slabs in fire 107 4.10 Design of connections under fire 109 4.11 Design of beam openings 109 4.12 Summary of structural fire design methods 110 4.12.1 Comparison of moment capacity method and critical temperature method 110 4.12.2 Comparison of three major methods 110 References 111

5 Typical fire safety design strategy for tall buildings 5.1 5.2

5.3

113

Introduction 113 Fire safety design objectives and strategies for tall buildings 113 5.2.1 Design objectives 114 5.2.2 Design strategies 114 5.2.3 Design process 114 Design strategy for tall buildings in fire 115 5.3.1 Prescriptive design 115 5.3.2 Performance-based design 115 5.3.2.1 Step 1: set fire safety goals and objectives 115 5.3.2.1 Step 2: determine performance criteria 115

Contents

5.3.2.2 5.3.2.3 5.3.2.4

5.4

5.5

5.6

5.7

5.8

5.9

xiii

Step 3: analysis of fire scenarios 116 Step 4: protection strategy 116 Step 5: determine whether the fire safety goals are met 116 5.3.3 Summary 117 Fire risk analysis for tall buildings 117 5.4.1 Qualitative fire risk assessment 117 5.4.2 Quantitative fire risk assessment 117 Deterministic and probabilistic assessments to determine the worst-case fire scenario 118 5.5.1 Deterministic approach 118 5.5.2 Probabilistic approach 118 Compartment design 119 5.6.1 Key components in a compartment 120 5.6.1.1 Fire doors design 121 5.6.1.2 Compartment wall design 122 5.6.1.3 Compartment floor design 122 5.6.2 Fire stop 122 5.6.3 Cavity barrier 123 5.6.4 Fire damper 124 5.6.5 Integrity of compartmentation in buildings 124 5.6.5.1 Measures to accommodate movements of compartment walls due to fire 124 5.6.5.2 Control movement of slab 125 Evacuation route design 125 5.7.1 Number of escapes routes and exits 126 5.7.2 Design of exits 126 5.7.3 Exit route 126 5.7.4 Travel distance 126 5.7.5 Staircases and elevators 128 5.7.5.1 Protected staircases and elevators 128 5.7.5.2 Fire lift lobby 128 5.7.5.3 External escape stairs 130 5.7.6 Phased/progressive evacuation 130 5.7.7 Refuge 130 5.7.8 Clear sign for evacuation 131 5.7.9 Computational models for evacuation simulation 131 Emergency vehicle and firefighter access 132 5.8.1 Equipment for firefighting 132 5.8.2 Firefighting lift, lobby, shaft, and stair 132 Resisting fire spread through building envelope 132

xiv

Contents

5.9.1 Resisting fire spread over external walls 133 5.9.2 Fire-resisting design for glazing 134 5.10 Fire detection, alarm, and communication system 134 5.10.1 Central fire alarm system 135 5.10.2 Smoke detections 135 5.10.3 Smoke control 136 5.11 Fire and smoke suppression system 136 5.12 Comparison for fire protection system for tall buildings across the world 137 5.13 Case study of fire safety deign for Burj Khalifa 137 5.13.1 Evacuation and refuge 141 5.13.2 Firefight access 142 5.13.3 Staircase and elevator 142 5.13.4 Alarm and warning system 142 5.13.5 Fire suppression 142 5.13.6 Special water supply and pumping system 142 5.14 Case study: structural fire design of the Shard 143 5.14.1 Introduction of the project 144 5.14.2 Structural system 144 5.14.3 Determine the worst-case fire scenarios 144 5.14.4 Design for fire resistance 145 5.15 Structural framing and structural system 146 References 147

6 Fire analysis and modeling 6.1 6.2

6.3

Introduction 149 Determining compartment fire 149 6.2.1 Simplified models from Eurocode 149 6.2.1.1 Compartment fires 150 6.2.1.2 Localized fires 150 6.2.2 Advanced models 150 6.2.2.1 Zone models 150 6.2.2.2 Limitations of zone modeling 154 6.2.2.3 Computational fluid dynamics (CFD) fire modeling 154 Determining member temperature 155 6.3.1 Simplified temperature increase models from Eurocode 155 6.3.2 Heat transfer using finite element method 156

149

Contents

6.3.2.1 6.3.2.2

Theoretical principles 156 Analysis software and modeling example 157 6.4 Determining structural response of structural members in fire 157 6.4.1 Multi-physics fire analysis (thermal mechanical coupled analysis) 158 6.4.1.1 Abaqus ® 158 6.4.1.2 ADINA 158 6.4.2 Sequentially coupled thermal-stress analysis 160 6.4.2.1 Sequentially coupled thermalstress analysis using Abaqus ® 160 6.4.2.2 Sequentially coupled thermal-stress analysis using ANSYS 161 6.4.2.3 Partial thermal-mechanical analyses in OpenSees 164 6.4.2.4 Codified thermal-mechanical coupled analysis 165 6.5 Probabilistic method for fire safety design 167 6.5.1 Reliability-based structural fire design and analysis 167 6.5.1.1 The basic reliability design principles 168 6.5.1.2 Reliability-based design and analysis procedure 169 6.5.1.3 Case study for reliability analysis for individual members 171 6.5.1.4 Case study for reliability analysis for a whole building 172 6.5.2 Fire fragility functions 174 6.5.2.3 Compartment-level fragility function 175 6.5.2.4 Building-level fragility function 175 6.5.3 Other probabilistic approaches in fire safety design 176 6.6 Major fire analysis software 177 6.6.1 Ozone software 177 6.6.2 CFAST 178 6.6.3 FDS 178 6.6.4 LS-DYNA 178 6.6.5 OpenSees 178 References 180

xv

xvi

Contents

7 Preventing fire-induced collapse of tall buildings 7.1 7.2 7.3 7.4

7.5

183

Introduction 183 Design objective and functional requirement for structural stability in fire 183 Importance of collapse prevention of tall buildings in fire 184 Collapse mechanism of tall buildings in fire 184 7.4.1 Factors affecting thermal response and failure mechanism of individual members 185 7.4.2 Behavior and failure mechanism of steel beams in fire 186 7.4.2.1 Local buckling of beams in connection area 186 7.4.2.2 Excessive deflection 187 7.4.3 Behavior and failure mechanism of slabs in fire 188 7.4.3.1 Membrane actions of slabs 188 7.4.3.2 Effect of different fire scenarios in composite slabs 189 7.4.3.3 Other research in composite slabs in fire 190 7.4.4 Behavior and failure mechanism of steel column in fire 191 7.4.4.1 Change of column force in fire 191 7.4.4.2 Out plane bending of columns 193 7.4.4.3 Effect of the slenderness ratios 194 7.4.5 Behavior of connections 194 7.4.6 Behavior and failure mechanism of concrete column in fire 195 Whole-building behavior of tall buildings in fire 195 7.5.1 Research of Fu (2016b) 195 7.5.2 Twin Tower collapse (WTC1 and WTC2) 195 7.5.2.1 Structural framing for WTC1 196 7.5.2.2 Reason for the collapse of WTC1 196 7.5.3 WTC7 197 7.5.3.1 Structural framing for WTC7 197 7.5.3.2 Reason for the collapse of WTC7 198 7.5.4 Cardington test 199 7.5.4.1 Severity of the fire 200 7.5.4.2 Structural framing 200 7.5.5 Other research in whole building behavior 201

Contents

xvii

7.6

Overall building stability system design for fire 202 7.6.1 Bracing system 202 7.6.2 Core wall design 202 7.7 Methods for mitigating collapse of buildings in fire 202 References 203

8 New technologies and machine learning in fire safety design 8.1 8.2

Introduction 205 New technologies in fire safety 205 8.2.1 PAVA alarm systems 205 8.2.2 IOT in fire safety 206 8.2.2.1 Fire safety sensors and BMS 206 8.2.2.2 Fire suppression 206 8.3 Machine learning in fire safety design 207 8.3.1 Machine learning and its application in the construction industry 208 8.3.2 Problems experienced in the conventional structural fire analysis approach 208 8.3.3 Predicting failure patterns of simple steel-framed buildings in fire 209 8.3.3.1 Define failure pattern 210 8.3.3.2 Dataset generation using the Monte Carlo simulation and random sampling 210 8.3.3.3 Training and testing 210 8.3.3.4 Failure pattern prediction 211 8.3.3.5 Fire safety design and progressive collapse potential check based on prediction results 211 8.3.4 Predicting and preventing fires with machine learning 211 8.3.5 Machine learning of fire hazard model simulations for use in probabilistic safety assessments at nuclear power plants 211 8.3.6 Learning algorithms and programming language 212 8.3.6.1 Learning algorithms 212 8.3.6.2 Programming language 212 References 212

205

xviii

Contents

9 Post-fire damage assessment

215

9.1 9.2

Introduction 215 Post-fire damage assessment 215 9.2.1 Post-fire damage assessment of concrete structure 215 9.2.1.1 Visual inspection 215 9.2.1.2 Schmidt rebound hammer 216 9.2.1.3 Petrographic analysis 216 9.2.1.4 Spectrophotometer investigations 216 9.2.1.5 Reinforcement sampling 217 9.2.1.6 Compression test 217 9.2.2 Post-fire damage assessment of structural steel members 218 9.2.2.1 Methods for post-fire damage assessment 218 9.2.2.2 Nondestructive post-fire damage assessment of structural steel members using the Leeb harness method 218 References 221

Index

223

Preface

I started to work on fire safety design for tall buildings since I joined WSP Group in 2007. I have been working on more than ten different types of buildings for fire safety design. I was very lucky to be one of the key team members for structural fire design of the currently tallest building in Western Europe, the Shard. To design fire safety for a building with such a complex nature is one of the most challenging jobs I have ever done. One of the major difficulties is that no systematic approach for structural fire design was available for tall buildings. Therefore, working together with my colleagues, we developed a systematic structural fire analysis framework for tall buildings via 3D finite element analysis in Abaqus®. The framework provided a cost-effective fire resistance design for the client. After joining the academia, I was invited by Institution of Structural Engineers to give lecture for practicing structural engineers on tall building design with one session to introduce fire safety design for tall buildings. While teaching, I noticed that most of the structural engineers lack fire safety design knowledge and are desperate to gain understanding of the relevant design and analysis methods. I have also been invited by Charted Institute of Architectural Technologists to give talks on fire safety design for tall buildings; in doing so, I found that most architectural engineers were also keen to gain the relevant knowledge. However, there are few textbooks available for fire safety design of tall buildings. Therefore, this textbook is designed to help fire safety engineering practitioners such as structural engineers, architectural engineers, and students to fully understand the principles of fire safety design and the relevant design guidance, particularly for tall buildings; explain effective ways to model different fire scenarios and analyze thermal response of building elements in fire; and introduce a systematic fire safety design approach for tall buildings.

xix

xx

Preface

Another feature of this book is that it demonstrates the 3D modeling techniques for fire safety analysis through the examples which replicate the real fire incidents such as Twin Tower, World Trade Center 7, and Cardington fire test. This would be helpful to the engineers in understanding the effective way to analyze the structures for fire safety design. Feng Fu

Acknowledgments

I would like to express my gratitude to Dassault Systems and/or its subsidiaries, ANSYS Inc. and/or its subsidiaries, and Autodesk Inc. to give me the permission to use the images of their product. I also thank BSI Group in U.K. and National Institute of Standards and Technology, Technology Administration, U.S. Department of Commerce for allowing me to reproduce some of the images from their reports. Some of the models, drawings, and charts used in this book are made by me and some are based on the work of my MSc and final-year students. I am very appreciative to my final-year and MSc students: Mr Shariq Naqvi, Mr Wing Sing Tsang, Mr Zubair Aziz, Mr Yu Zhang, and Mr Xiao Li. I also thank Mr Xuandong Chen for his help. I am thankful to all reviewers who offered their comments. Special thanks to Tony Moore and Frazer Merritt from Taylor & Francis for their assistance in preparation of this book. Thanks to my family, especially my father Changbin Fu, my mother Shuzhen Chen, and my wife Dr. Yan Tan for their support in finishing this book.

xxi

Author

Dr. Feng Fu, Ph.D., MBA, CEng, FIStructE, FASCE, FICE, FHEA, worked for several world-leading consultancy companies including WSP Group, where he was one of the key team members in structural fire design of the tallest building in Western Europe, the Shard. Currently, he serves for two building design standard committees of the American Society of Civil Engineers and also acts as an associate editor and editorial board member for three international journals. He has published more than 100 technical papers and three textbooks including Structural Analysis and Design to Prevent Disproportionate Collapse (CRC Press, 2016).

xxiii

Chapter 1

Introduction

1.1 AIMS AND SCOPE Across the world, there are many fire incidents happening in tall or multistorey buildings every year. It causes loss of life and damage to the properties. About 69% fires are caused by electricity. As a result, fire safety is one of the key tasks in tall building designs. Particularly, a large percentage of tall buildings are steel structures or partially made of steel frames which require more stringent fire safety design requirements. The recent disaster in Grenfell Tower (Fu, 2017) caused huge casualties. It has embarked increasing concerns from the public and building engineers in fire safety design for tall buildings. Fire development and subsequent thermal response of the building depend upon numerous factors, invariably featuring a high degree of uncertainty. While permitted within performance-based frameworks and supported by design codes (EN 1991-1-2, 2002; EN1992-1-2, 2004; EN 1993-1-2, 2005; EN 1994-1-2, 2005), the appraisal of structural response in fire is challenging given the sources of uncertainty that exist. This is primarily due to the complexity caused by different fire scenarios which can possibly be formed when fire occurs. In addition, for a structure such as a tall building, the structural systems are much more complicated, which also brings extra difficulties in the structural fire analysis. In the past two decades, an increasing number of tall buildings have been built worldwide. Advancements in structural engineering make possible the increase in height, size, and complexity of modern tall buildings. Particularly, in modern tall building design, the structural system of tall buildings becomes increasingly complicated. Figure 1.1 shows the structural system of a newly built tall building in Beijing with a height of 528 m. The lateral stability system comprises so-called Mega Frame system and Diagrid system. Its more complex structural system makes its fire safety design a challenging job. In order to effectively design fire safety for tall buildings, it is essential to understand the behavior of the buildings in fire. In addition, a significant amount of new construction materials— elements such as new type of cladding systems and new construction 1

2

Fire Safety Design for Tall Buildings

Figure 1.1 China Zun Tower in construction. (Photo taken by the author’s father.)

techniques—have also been developed. The tall building of today is different from that of a decade ago with foreseen changes even greater in the immediate future. These advancements make fire safety design for tall buildings an even more challenging task for design engineers. In the current design practice, detailed fire safety design guidelines has been developed across the world, such as Eurocode (EN 1991-1-2, 2002; EN1992-1-2, 2004; EN 1993-1-2, 2005; EN 1994-1-2, 2005). As a design engineer, it is imperative to guarantee that in the design process, sufficient measures for fire safety should be made. An engineer should also have the capacity to analyze the response of structures under different fire scenarios and subsequent fire protection measures using appropriate procedure and analysis software. Therefore, this textbook is designed to help fire safety engineering practitioners such as structural engineers, architectural engineers, and

Introduction

3

students to fully understand the principles of fire safety design and the relevant design guidance, particularly for tall buildings; the effective way to model different fire scenarios and thermal response of building in fire; and introduction of a systematic fire safety design approach for tall buildings. Detailed demonstrations of 3D modeling techniques for fire safety analysis are also made. In addition, case studies based on various fire scenarios and different structural layouts of tall buildings are provided to demonstrate failure mechanisms of buildings in fire and effective design methods for fire safety. 1.2 MAIN FIRE SAFETY DESIGN ISSUES FOR TALL BUILDINGS The main objective of the fire safety design for tall buildings is life safety of the occupants. Therefore, all the design processes are centered around life safety. Among them, compartmentation design, evacuation route design, and structural fire design are the three key design focuses. These three factors are affecting each other, for example, when designing the evacuation route, the time of evacuation is affected by the time of failure of structural members. The size and layout of the compartment also affect the evacuation route design. The integrity of the compartment is very important in containing the fire in its original place or delaying its spread. However, the integrity of the compartment is greatly affected by structural fire design. For example, the deformation of the compartment wall in fire reduces its capacity to maintain its integrity. Its deflection is controlled primarily by its supporting beams. Designing a beam with reduced deflection in fire will improve the whole integrity of the compartment. These design processes will be explained in detail in Chapters 3–5. Fire scenario analysis and its corresponding structural fire analysis are both unique and complicated procedures, and they also require an engineer to have the ability to use modern commercial software for fire scenarios simulation or a finite element package to analyze the structural responses of a building in fire. Therefore, this book also features a detailed introduction to the use of fire analysis software such as OZONE and FDS®1 as well as finite element programs such as Abaqus®, 2 ANSYS, and LS_DYNA OpenSees, ADINA. 1.3 STRUCTURE OF THE BOOK Chapter 1 is the introduction of the book. It introduces the aims and scope, as well as the structure, of this book. Chapter 2 introduces several fire incidents that happened in tall buildings, followed by the regulatory requirements from various codes across the

4

Fire Safety Design for Tall Buildings

world. At the end of this chapter, the basic principles for fire safety design of tall buildings will be discussed. Chapter 3 introduces the fundamental knowledge of fire and fire safety design. The characteristics of fire and its development are introduced at the beginning. The key fire scenarios that affect the performance of the building members in fire—such as ventilation-controlled or fuel-controlled fire and long-cool, short-hot fire—will be explained. In addition, the fundamentals of heat transfer, a process of the heating up of structural members due to fire, will be introduced. The basic structural fire design principles will also be explained. In fire safety design, most of the codes specify the fire resistance of building elements. The relevant information will be provided in the latter part of this chapter followed by the introduction of fire protection methods. Chapter 4 introduces the structural fire design in depth on the basis of Chapter 3. It introduces the structural fire design procedures for steel, concrete, and composite structural members based on Eurocodes and British Standards. Two key structural fire design methods are introduced: critical temperature method and moment capacity method. It also covers the design of post-tensioning slabs, connection, and beams with openings. Chapter 5 provides a detailed design strategy for tall buildings. The prescriptive and performance-based fire design approaches are first introduced, followed by fire risk analysis. The deterministic and probabilistic approach to determine the worst-case fire scenarios is then introduced. A detailed demonstration of compartment design and evacuation route design for tall buildings is made followed by the description of other design issues such as firefighter access and fire protection requirement to façade. The fire alarm system, communication system, fire and smoke suppression system are also discussed. At the end of this chapter, case studies for two real construction projects, namely, Burj Khalifa and the Shard, are made. Chapter 6 introduces various theoretical and numerical methods for fire analysis. It starts with the method to determinate the compartment fire including a detailed introduction of Zone model and CFD model. It is followed by the methods of solving thermal response of structural members such as heat transfer analysis and thermal–mechanical analysis. In addition, the probabilistic method for fire safety analysis will be covered. In the final part of this chapter, various numerical modeling software for fire analysis will be explained. Chapter 7 discusses how to design a building to prevent fire-induced collapse. The collapse mechanism of a tall building in fire and methods for mitigating the collapse are introduced, all based on existing research and fire-induced collapse incidents. Chapter 8 introduces new technologies developed for fire safety design, such as PAVA system, IOT, and smart building management system. Some

Introduction

5

pilot studies of using machine leaning in fire safety design will also be introduced in this chapter. Chapter 9 introduces the post fire damage assessment methods. Different damage assessment techniques including destructive and nondestructive assessment methods for concrete and steel structures are introduced.

NOTES 1 FDS https://www.nist.gov/services-resources/software/fds-and-smokeview. 2 Abaqus is a registered trademark of Dassault Systemes S.E. and Its affiliates.

REFERENCES EN 1991-1-2 (2002), Eurocode 1. Actions on Structures-Part 1-2: General actions. Actions on structures exposed to fire. Commission of the European communities. EN1992-1-2 (2004), Eurocode 2. Design of concrete structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. EN 1993-1-2 (2005), Eurocode 3. Design of steel structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. EN 1994-1-2 (2005), Eurocode 4. Design of composite steel and concrete structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. Fu, F. (2017), Grenfell Tower disaster: How did the fire spread so quickly? BBC Australia.

Chapter 2

Regulatory requirements and basic fire safety design principles

2.1 INTRODUCTION In this chapter, several fire incidents that happened in tall buildings will be first introduced followed by the regulatory requirements from various codes across the world. At the end of this chapter, the basic principles for fire safety design of tall buildings will be discussed.

2.2 FIRE INCIDENTS AND FIRE TESTS OF TALL BUILDINGS WORLDWIDE Table 2.1 shows the fire incidents occurred each year in different countries during 2012–2015. It can be seen that there are a large number of fire incidents happening each year across the world. Table 2.1 Fire incidents across the world from 2012 to 2015 in different countries No. 1 2 3 4 5 6 7 8 9

Country US Bangladesh Russia Japan Vietnam Germany France Great Britain Italy

Population in 1,000 323,128 154,331 146,270 128,130 93,000 82,218 66,628 63,786 61,000

2012

2013

2014

2015

1,375,000 1,240,000 1,298,000 1,345,500 17,504 17,912 17,830 17,488 162,900 152,959 150,437 145,900 44,101 48,095 43,741 39,111 1,900 2,540 2,375 2,451 175,354 192,078 306,871 281,908 270,900 300,667 272,800 192,700 212,500 191,647 241,232 196,196 189,375 234,675 (Continued )

7

8

Fire Safety Design for Tall Buildings

Table 2.1 (Continued) Fire incidents across the world from 2012 to 2015 in different countries No.

Country

Population in 1,000

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Myanmar Spain Ukraine Poland Canada Malaysia Peru Nepal Taiwan Romania Kazakhstan Netherlands Greece Belgium Czech Republic Sweden Hungary Jordan Belarus Austria Switzerland

51,486 47,079 42,673 38,454 35,544 31,800 30,741 30,430 23,069 20,121 17,500 16,979 10,788 10,700 10,579 9,851 9,830 9,700 9,505 8,740 8,372

2012

2013

1,219 1,673 142,500 135,000 71,443 61,144 183,888 125,425 45,005 37,194 29,874 33,640 11,329 11,264 1,021 1,574 1,451 38,077 16,145 13,621 33,731 28,232 21,369 21,228 20,492 16,563 22,657 25,392 37,106 20,177 25,644 23,961 34,505 7,151 42,213 40,395 14,304 12,893

2014 1,629 128,000 68,879 145,237 36,445 54,540 9,430 958 1,417 14,477 91,160 17,388 19,536 20,795 7,489 43,336 11,658

2015 137,000 79,640 184,847 40,865 9,473 1,704 26,247 14,452 125,200 20,232 22,785 21,056 32,488 7,339 45,349 12,477

Source: Fi. (https://www.ctif.org/sites/default/files/2018-06/CTIF_Report23_World_Fire_ Statistics_2018_vs_2_0.pdf)

Most of the fire will cause local damages to the buildings; however, some may even causes collapse the entire buildings. The two famous examples of fire-induced building collapse are Twin Towers and WTC7. They will be introduced in detail in this chapter.

2.2.1 Grenfell Tower Grenfell Tower fire (Fu, 2017) caused a great tragedy with 71 deaths and over 70 injuries. Among the 129 flats, occupants of 23 flats died and 223 people escaped. There are a number of factors in the design of the 24-storey tower that may have contributed to the speed and scale of the blaze. The fire started in a kitchen at lower level of the tower, and then the flame propagated through the cladding to the upper level of the tower, causing the fire to spread to almost the entire building (Figure 2.1).

Regulatory requirements & design principles

9

Figure 2.1 Fire incidents in Grenfell Tower. (This file is licensed under the Creative Commons Attribution 4.0 International license, https://commons.wikimedia. org/wiki/File:Grenfell_Tower_fire_(wider_view).jpg.)

2.2.1.1 The new cladding system As shown in Figures 2.2 and 2.3, it was reported that a new cladding was added 1 year before the fire. From Figure 2.4, it can be seen that there are three layers in the cladding. The material used for the cladding was primarily aluminum, with an extra layer of insulation in between the aluminum layers. British Research Establishment conducted a detailed investigation after the fire (British Research Establishment, 2017). Both the cladding panels and the infilled insulation of the new façade were tested under fire. It was found that they were not good in terms of fire resistance. What’s more, aluminum has high conductivity, so the cladding itself could have heated up very quickly, failing to prevent the fire from spreading through the windows and up the exterior of the block from one storey to another. 2.2.1.2 Compartment and evacuation route for Grenfell Tower Most of the current guidelines across the world contain detailed design requirements for fire safety. But at the time Grenfell Tower was built (in 1974),

10

Fire Safety Design for Tall Buildings

Figure 2.2 Grenfell Tower before refurbishment. (This file is licensed under the Creative Commons Attribution-Share Alike 2.0 Generic license, Attribution: Robin Sones, https://upload.wikimedia.org/wikipedia/commons/b/b4/Grenfell_Tower%2C_ London_in_2009.jpg.)

the rules and regulations were not as stringent as now; therefore, most of the old buildings did not conform to the latest guidelines for fire safety design. Hence, it is imperative to update them by making installation of sprinklers, fire alarms, and extra fire evacuation staircases mandatory. 2.2.1.2.1 Compartmentation One of the key strategies in fire safety design is to correctly design fire compartments to keep the fire from spreading quickly. To contain the fire in a

Regulatory requirements & design principles

11

Figure 2.3 New cladding burnt by fire. (This file is licensed under the Creative Commons Attribution 4.0 International license, https://upload.wikimedia.org/wikipedia/ commons/9/9a/Grenfell_Tower_fire_morning.jpg.)

Alum

inium

Concrete Insula tion Cla

ddi

ng

sheet

Cor

Air

e Alu min ium

cap

acit

y

Figure 2.4 Schematic layout of the new cladding.

she et

12

Fire Safety Design for Tall Buildings

local area, placing barriers in the building—such as fire-resistant doors (fire doors) and walls (compartment walls)—is essential. These design measures can at least slow down the speed at which it spreads. These compartments are designed by architects based on the function of the buildings, so residential and commercial buildings will have different compartment design strategies. The report of British Research Establishment (2017) has also noticed this. It is also reported that, ironically, fire doors were only installed in the storage rooms; therefore, all the stuff inside the storage room remained intact during the fire. No fire doors were installed in any flats. Figure 2.5 shows the structural layout of the Grenfell Tower, and no compartment wall and fire door were used apart from the fire door in the storage room in the entire floor, so the whole floor can be treated as one compartment, which means that when fire starts at one room, it will quickly spread into other rooms. In the current design practice, some buildings even include special design measures for fires, such as refuge rooms in higher storeys for occupants who could have trouble escaping downstairs. There are also active fire protection methods such as using sprinklers. No sprinkler seems to be installed in Grenfell Tower. A local residents action group also claimed that their warnings about a lack of fire safety measures “fell on deaf ears.”

1

2

3

4

A

Apartment 1

Apartment 2

B

Apartment 3

Apartment 4

C

Apartment 5

D

Figure 2.5 Compartment layout of Grenfell Tower.

Apartment 6

5

Regulatory requirements & design principles

13

2.2.1.2.2 Evacuation The evacuation route is another most important design element in fire safety. For tall buildings, staircase is the major evacuation route for the occupants. The route should allow occupants to escape the building as quickly as possible while sheltering them from smoke and flames. Approved Document B (2019) requires that every storey with a floor level more than 11 m above ground level has an alternative means of escape. Some tall buildings have staircases installed on the outside to prevent people from getting stuck in the corridors and provide access to fresh air while they escape. As it can be seen from Figure 2.6, there was only one set of stairs for evacuation in Grenfell Tower. It is common for these kinds of old tower

D

C

B

Figure 2.6 Evacuation route of Grenfell Tower.

A

14

Fire Safety Design for Tall Buildings

blocks to have only one staircase, and no extra backup staircases available. Therefore, it slowed down the speed of the evacuation. It’s clear that residents were not happy with the fire safety of the escape route from a blog posted before the disaster. 2.2.1.3 Collapse potential for Grenfell Tower According to reports, Grenfell Tower is a concrete-framed block, which has high rate of fire resistance. While steel structural members can buckle in high temperatures, concrete structures can help to prevent the collapse of a building in case of fire, as well as making it safer to use helicopters— which can dump up to 9,842 litres of water at a time—to extinguish the blaze. 2.2.1.4 Major findings from Interim Report of British Research Establishment (2017) The following major findings are reported from the report of British Research Establishment (2017): 1. 2. 3. 4.

The aluminum panels and insulation were used in the façade. There was a compartmentation problem, i.e., lack of fire doors. The new windows were too narrow. The cavity barriers were of insufficient size.

Findings 1 and 2 have been discussed in the previous section. For finding 3, the UK regulation Approved Document B: Volume 1 (2019) makes different recommendations corresponding to the height of the building: a window might be an appropriate means of escape from a flat located on an upper level of a building (no higher than 4.5 m from ground level), but flats at a higher elevation would require alternative forms of emergency exits. Officially, as an old building, Grenfell Tower did not follow this regulation. For finding 4, a cavity barrier can be a roof void barrier, underfloor cavity barrier suited to IT suites and offices with raised access flooring and an edge of slab fire protection detail between the building facade and the floor slab typically found in high-rise residential buildings. Cavity barrier is a vital fire seal of buildings, which in many cases is unseen but plays a vital role in containing fire and smoke within cavities at 20 m divisions. Approved Document B (2019) is a fire safety regulation issued by the UK government, which will be introduced in the next section. From the investigation, questions have been raised about the state of fire safety regulation

Regulatory requirements & design principles

15

in England and Wales, particularly for high-rise buildings. Following allegations that the cladding used on Grenfell Tower may have contributed to the rapid spread of the fire, more stringent regulations were introduced in the new version of Approved Document B (2019), which will be introduced in Chapter 5.

2.2.2 Twin Tower It is well known that on September 11, 2001, the Twin Towers (referred to as WTC1 and WTC2, respectively) were collapsed due to aircrafts crashing. However, the aftermath investigation by National Institute of Standards and Technology (NIST, 2005) shows that the collapse of the buildings is not due to the impact load from the aircrafts but due to the fires. As shown in Figure 2.7 and Figure 2.8, the towers of WTC1 and WTC2 were designed with closely spaced steel mega columns at perimeter and steel cores in the center to provide robust stability system for such tall buildings. In between the perimeter and the center, there is a large column-free space which was bridged by prefabricated floor trusses. The whole structural system of the buildings can be simplified as shown in Figure 2.9. Apart from this so-called “Tube in Tube” system, the towers also used the conventional outrigger truss (also called Hat Truss) between the 107th and 110th floors

Figure 2.7 Structural system of WTC1 and WTC2. (Usrlman, the copyright holder of this work, release this work into the public domain, https://commons.wikimedia.org/wiki/File:WTC_bathtub_east.JPG.)

16

Fire Safety Design for Tall Buildings

101103 106 109 112 115 118 121 124 127 130 133 136 139 142 145 148 151 154 157 159 100

200

459 457

201 203

454

206

451

209

448

212 215

445 442

501

502

503

601

602

603

504 505

506

604 605

606

507

218

508

221

439 436

607

608 224 227

433 430 427

701

702

801

802

703 803

704 705

706

804 805

707

708

806

807

230 233 236

424 421 418

901

902

903

904 905

906

1001

1002

1003 10041005

1006

907

908

1007

1008

239 242

415

245

412

248

409

251

406

254

403 401

257 259 300

400 359357 354 351 348 345 342 339 336 333 330 327 324 321 318 315 312 309 306 303 301

Figure 2.8 Typical floor layout for World Trade Center 1. (Reprinted from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center Disaster: Final report on the collapse of the World Trade Center Towers (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1, Figure 1–3, p. 7. https://doi.org/10.6028/ NIST.NCSTAR.1.)

to further strengthen the cores. These special structural configurations guarantee a very strong structural system to resist terrific lateral load such as wind or earthquake. Therefore, it is fully capable to resist the impact load caused by the crashing of the aircrafts. For fire safety reason, the whole steel frame system including steel core and perimeter columns was also protected with sprayed-on fire-resistant material which can protect the structural members from fire. However, disengagement of the fire protections was reported by NIST (2005). The report from NIST (2005) shows that after the aircrafts with full tanks of fuel hit WTC1 and WTC2, the explosion caused fire inside the two buildings. When the fire protection ceased to protect the structural members from the fire, the fire caused the sagging of the floor trusses, which further caused inward pulling of the perimeter columns. “This led to the inward bowing of

Regulatory requirements & design principles

17

Perimeter Steel Box Column

Composite truss floor system

Inner Steel Core

Figure 2.9 Schematic drawing of the structural system of Twin Towers.

the perimeter columns and failure of the south face of WTC1 and the east face of WTC2, initiating the collapse of each of the towers” (NIST, 2005). A 2D model was built by Usmani et al. (2003) to simulate the collapse mechanism of the Twin Tower, as shown in Figure 2.10. It can be seen that the bowing of the floor caused buckling and failure of the perimeter columns as indicated in NIST (2015). It also clearly shows the progression of the failure at different times. As time passed by, the bowing of the floor and subsequent failure of the columns caused the overall failure of several storeys. When these storeys gave up, the storeys above them started to free-fall. It further increased the load for the storeys at the lower level, thus causing a so-called progressive collapse as shown in Figure 2.11. t=2000

t=2500

t=3600

Figure 2.10 Collapse mechanism of Twin Towers (Usmani et al., 2003).

18

Fire Safety Design for Tall Buildings

Figure 2.11 The process of progressive collapse in WTC1.

2.2.3 World Trade Center 7 World Trade Center building 7 is another tall building that collapsed due to fire. It is a 47-storey commercial building located next to the Twin Towers. As shown in Figure 2.12, the WTC7 is in irregular trapezoid shape. It is a steel composite-framed building. The main lateral stability is through steel moment connection frames with a center rectangular building core comprising 21 steel internal columns. According to the investigation report of NIST (2008), the collapse of WTC7 was mainly due to the fire ignited as a result of the debris from the collapse of WTC1. There were both passive and active fire protection systems in WTC7 (NIST, 2008); the passive fire protection system is the sprayed fire-resistive material (SFRM) on the structural steel and metal decking for the floors. The active fire protection system consists of sprinklers inside the buildings; however, it was not functioning as the main water supply was cut off during the accident (NIST, 2008). The fire started to propagate inside the building, causing the buckling of columns, resulting in the failure of the floor above and further failure of the adjacent columns in the horizontal direction which triggered the progressive collapse of the whole buildings.

2.2.4 Other fire incidents of tall buildings 2.2.4.1 First Interstate Bank building in Los Angles This is a 62-storey building using steel composite structural system. Fire started on level 12 and spread to four floors above. The building’s steel beams and columns were fire protected; however, no fire compartment was designed. Therefore, there were no compartment walls and compartment floors. let alone fire stops. This caused nearly five storeys to burn out and partial building collapse.

Regulatory requirements & design principles

19

Figure 2.12 WTC7 in fire.(http://upload.wikimedia.org/wikipedia/commons/thumb/0/0e/ Wtc7onfire.jpg/511px-Wtc7onfire.jpg, Courtesy of the Prints and Photographs Division. Library of Congress, public domain.)

2.2.4.2 Plasco shopping center, Iran It is a 17-storey residential and commercial building using steel composite structural system, but no sprinkler or any other fire protection measures were adopted. Fire started on level 10 and spread to upper floors, causing the entire building to collapse (BBC, 2017). 2.2.4.3 Faculty of Architecture Building, Delft University The fire started in a coffee-vending machine on the sixth floor of the 13-storey Faculty of Architecture Building at the Delft University of Technology, Delft, the Netherlands. The building had a reinforced concrete and steel structure consisting of a combination of six, three-storey structures, which effectively served as a podium, with a 13-storey tower located above. Although all occupants of the building were evacuated safely, the

20

Fire Safety Design for Tall Buildings

fire spread rapidly which severely impacted the operations of fire department, and thus the fire burned uncontrollably for several hours, eventually resulting in the structural collapse of a major portion of the building. The investigation of Meacham et al. (2010) and Engelhardt et al. (2013) shows the possibility of fire-induced spalling in RC columns leading to structural collapse due to excessive deflection of slabs. This is a very rare example of concrete-building collapse due to fire. As it is widely known, the concrete is good in fire resistance. However, the fire burned for several hours before its extinction, which made the concrete to spall. The spalling can further induce the failure of the reinforcement, which will be explained in detail in Chapter 4. 2.2.4.4 Windsor Tower, Spain It is a 32-storey building having a steel frame with concrete core wall as the major lateral stability system. Fire started on level 12 during the construction stage, causing partial collapse of the building. As the building was in the construction stage when the fire started, the sprinkler system and fire protection for exterior columns were not completely finished (Scoss Failure Data Sheet, 2008).

2.2.5 Cardington fire test The Cardington fire test (British Research Establishment, 1999) is the first full-scale fire test for a multistorey building in the history. Fire tests were performed on an eight-storey typical braced steel office building at Cardington in the UK. As shown in Figure 2.13, a real Range Rover car was parked beside the building as an indication of the real size of the building tested in fire. The Cardington fire test is a milestone in fire safety design, as it clearly discovered the failure modes of structural components in case of fire, and it laid a foundation for modern structural fire design. 2.2.5.1 Introduction of the test Figure 2.14 is a plan layout indicating the locations of different fire tests performed. Altogether six tests representing different scenarios were performed. 2.2.5.1.1 Test 1—single secondary beam— gas-fired furnace As shown in Figure 2.14, Test 1 was a single secondary beam test. The beam and supported slab were heated up to 800°C–900°C using a gas furnace with the connections remaining at ambient temperature.

Regulatory requirements & design principles

21

Figure 2.13 Full-scale multistorey building used for Cardington tests (British Research Establishment, 1999).

2.2.5.1.2 Test 2—plane frame test—gas-fired furnace As shown in Figure 2.14, Test 2 was designed to investigate primary beams and columns along gridline B which supported the fourth floor. The primary and secondary beams and top columns were left unprotected. They were heated using gas furnace.

22

Fire Safety Design for Tall Buildings

A

B 9.000m

C 9.000m

D 9.000m

E 9.000m

4 6.000m

F 9.000m

Test 4

3 9.000m

Test 2

2 6.000m

Test 1 Test 3

1

Figure 2.14 Floor layout and location of Cardington fire tests (British Research Establishment, 1999).

2.2.5.1.3 Tests 3 and 4—corner test—timber cribs = 45 kg/m2 As shown in Figure 2.14, both Tests 3 and 4 are corner compartment tests to investigate the composite floor system and membrane effect. They were heated up using wood cribs. It is equivalent to 45 kg/m 2 wood in fire. In Test 3, all structural members were left unprotected apart from columns, column-to-beam connections, and external perimeter beams. The maximum recorded steel temperature was 935°C. In Test 4, only columns were protected. Windows and doors were closed to simulate the fuel-controlled fire development, as the level of oxygen was restricted. It was recorded that the maximum steel temperature was 903°C. 2.2.5.1.4 Test 5—large compartment test Test 5 was a large compartment test (340 m 2) which was conducted between the second and third floors. A fire resistance wall was constructed along the full width of the building. All the steel beams were left unprotected. The maximum atmosphere temperature and steel temperature recorded were 746°C and 691°C, respectively. Many beam-to-beam connections were found to have locally buckled, and many end-plate connections fractured down one side after cooling. 2.2.5.1.5 Test 6—demonstration test—office furniture ~45 kg/m2 wood Test 6 is to simulate the fire that occurs in an office; to closely simulate office environment, old office furniture was used as fuel in an open compartment of 135 m 2 . It is equivalent to 45 kg/m 2 wood in fire. Only columns and beam-to-column connections were protected. The maximum steel temperature was 1,150°C.

Regulatory requirements & design principles

23

2.2.5.2 Failure modes for buildings in fire The Cardington fire test helps us to fully understand the failure modes of a multistorey building in fire, which provides guidance for structural fire design of tall buildings. Four major types of failure modes were observed during the full-scale fire tests. They are introduced in the following. 2.2.5.2.1 Beam buckling and yielding As can be seen from Figure 2.15, during Test 1, local buckling at lower flange at the ends of the beam due to the restraining forces occurred due to thermal expansion against the web of the column section. Yielding at both ends of the test beam was also observed during the experiment. In Test 4, distortional buckling was observed along most of the beam length. 2.2.5.2.2 Column buckling and yielding As shown in Figure 2.16, during Test 2, it is observed that the exposed parts of the columns squashed at approximately 670°C. This may lead to local collapse. Therefore, it is suggested by British Research Establishment (1999, 2004) that the columns should be fully fire protected along the entire length. In Test 3, extensive buckling was noticed at beam-to-column connections. The end of an internal secondary beam which was connected to a primary beam buckled locally due to axial restraint from adjacent members. However, no local buckling occurred at the other end of the beam

Figure 2.15 Beam buckling (British Research Establishment, 1999).

24

Fire Safety Design for Tall Buildings

Figure 2.16 Column buckling and yielding (British Research Establishment, 1999).

which was connected to an external beam. This was because the thermal expansion of the secondary beam caused the external beam to twist, resulting in insufficient restraint to cause local buckling. 2.2.5.2.3 Connection failure In Test 3, it is also noticed that the bolts at the fin-plates connection were sheared off as shown in Figure 2.17.

Figure 2.17 Connection failure (British Research Establishment, 1999).

Regulatory requirements & design principles

25

Figure 2.18 Slab failure (British Research Establishment, 1999).

2.2.5.2.4 Slab failure During Test 6, it is found that no signs of failure were observed, but there were extensive cracking formed near the column zone during the latter phase of cooling as shown in Figure 2.18.

2.2.6 Discussion From the introduction of the above incidents and Cardington fire tests, it can be seen that for the two steel framed buildings, WTC1 and WTC7 both collapsed. However, the concrete building Grenfell Tower did not. Though the concrete building in Delft University of Technology collapsed due to fire, it is quite rare. It can be seen that due to its fire resistance feature, concrete is the best option in the fire safety design. However, a pure concrete structure is not feasible. Most of the buildings are still steel framed. Therefore, fully understanding the design strategy and design method of tall buildings in fire is essential for designers.

26

Fire Safety Design for Tall Buildings

In a tall building design, the staircase situated in the core of the buildings is one of the major evacuation routes for the occupants, which should be kept free of flame and smoke in the event of fire. Therefore, after the 9/11 incident, almost all the newly built tall buildings use concrete as the major material of core, and steel cores are not used for tall buildings.

2.3 CURRENT DESIGN GUIDANCE AND REGULATIONS TO FIRE SAFETY IN HIGH-RISE BUILDINGS In the current design practice, there are several design codes and regulations for fire safety design used worldwide. A brief introduction of this guidance will be made in this section.

2.3.1 British design guidance and regulations 2.3.1.1 Building Regulations 2010—Approved Document B The British government issued a series of approved documents that give practical guidance about how to meet the requirements of the Building Regulations 2010 for England. Approved Document B (2019) is a particular building regulation document that provides guidance for fire safeties of common buildings. It set out minimum standards in respect of health and safety which must be met when constructing new buildings. The fire design regulations in the UK take a dual approach: imposing requirements on new buildings (or those which undergo material alterations) and a separate set of requirements for existing buildings. In the case of existing multiple-occupancy buildings, obligations to ensure fire safety are imposed on both the “responsible persons” for a building, which can encompass owners, landlords, and managing agents, and also social housing landlords such as local authorities, housing associations, and any managing agents or tenants’ association with delegated responsibilities. It comprises five major parts as follows. 2.3.1.1.1 Requirement B1: means of warning and escape This part stipulates the regulations on fire detection and alarm system and the means of escapes from different levels of the buildings.

Requirement “Means of warning and escape B1. The building shall be designed and constructed so that there are appropriate provisions for the early warning of fire, and appropriate

Regulatory requirements & design principles

27

means of escape in case of fire from the building to a place of safety outside the building capable of being safely and effectively used at all material times.” 2.3.1.1.2 Requirement B2: internal fire spread—linings This part stipulates the regulations on a restricted spread of fire over internal linings. The building fabric should make a limited contribution to fire growth, including a low rate of heat release.

Requirement: “Internal fire spread (linings) B2. (1) To inhibit the spread of fire within the building, the internal linings shall— (a) adequately resist the spread of flame over their surfaces; and (b) have, if ignited, either a rate of heat release or a rate of fire growth, which is reasonable in the circumstances. (2) In this paragraph ‘internal linings’ means the materials or products used in lining any partition, wall, ceiling or other internal structure.” 2.3.1.1.3 Requirement B3: internal fire spread—structure This part stipulates the regulations on the minimum fire resistance for the load-bearing structural elements such as frames, beams, columns, and loadbearing walls. It also stipulates on the design of the compartment.

Requirement “Internal fire spread (structure) B3. (1) The building shall be designed and constructed so that, in the event of fire, its stability will be maintained for a reasonable period (2) A wall common to two or more buildings shall be designed and constructed so that it adequately resists the spread of fire between those buildings. For the purposes of this sub-paragraph a house in a terrace and a semi-detached house are each to be treated as a separate building. (3) Where reasonably necessary to inhibit the spread of fire within the building, measures shall be taken, to an extent appropriate to the size and intended use of the building, comprising either or both of the following— (a) sub-division of the building with fire-resisting construction. (b) installation of suitable automatic fire suppression systems. (4) The building shall be designed and constructed so that the unseen spread of fire and smoke within concealed spaces in its structure and fabric is inhibited.”

28

Fire Safety Design for Tall Buildings

2.3.1.1.4 Requirement B4: external fire spread This part stipulates the regulations on resisting fire spread over external walls.

Requirement “External fire spread B4. (1) The external walls of the building shall adequately resist the spread of fire over the walls and from one building to another, having regard to the height, use and position of the building. (2) The roof of the building shall adequately resist the spread of fire over the roof and from one building to another, having regard to the use and position of the building.” 2.3.1.1.5 Requirement B5: access and facilities for the fire service This part stipulates the regulations on access and facilities for the fire service.

Requirement “Access and facilities for the fire service B5. (1) The building shall be designed and constructed so as to provide reasonable facilities to assist fire fighters in the protection of life. (2) Reasonable provision shall be made within the site of the building to enable fire appliances to gain access to the building.” 2.3.1.1.6 Summary It is important to note that the UK Building Regulations do not specify how these standards should be met, but how compliance achieved. There is no legal requirement to implement the guidance, provided that the minimum standards are met. Approved Document B, which has separate volumes for dwelling houses (Volume 1) and for other buildings (Volume 2). Part B also requires that buildings be constructed in a manner to limit both internal and external fire spreads. Approved Document B provides prescriptive examples of how this can be achieved, but as noted above, compliance with Approved Document B is not mandatory. The penalty for contravention of the Building Regulations is an unlimited fine. 2.3.1.2 The FSO and Housing Act 2004 For existing buildings, there are other two regulations in UK which can be referred to: one is Fire Safety Order 2005 (FSO), which imposes duties on individuals in control of the building, the other one is Housing Act 2004 (the Act), which imposes monitoring duties on local authorities to take enforcement action against those in control of the building.

Regulatory requirements & design principles

29

2.3.1.2.1 FSO FSO imposes a duty on the “responsible person” to implement fire safety measures including undertaking a risk assessment; making fire safety arrangements; ensuring that a premise is equipped with fire-fighting equipment, fire-detection equipment, and emergency routes; and establishing procedures such as fire safety drills a “responsible person” will be a freehold owner or landlord but may also include a “residential management company” or a “right to manage company.” Fire risk assessments should be conducted on a periodic basis to evaluate the risk and determine the appropriate fire safety measures to be implemented and maintained. For tall buildings, the FSO only applies to the common parts. Responsible persons must ensure that tenant activities do not compromise the safety of the common parts (e.g. by placing obstructions/flammable objects in corridors or blocking fire-escape routes). Landlords should take appropriate action to minimize such risks, for example, by placing signs in prominent places instructing tenants to keep the common parts free from any obstruction and/or flammable objects. Landlords need to be mindful of these risks when drafting tenancy repair and maintenance obligations and should also ensure that regular checks are carried out/safety audits conducted on a routine basis. Certain repair works, including the replacement of fire doors, may constitute “material alterations” and must also comply with the Building Regulations. Failure to provide adequate fire safety measures is a criminal offence. If the failure places one or more people at risk of death or serious injury, it can be punishable by an unlimited fine and/or up to 2 years’ imprisonment. The Fire and Rescue Authorities have the responsibility for the enforcement of the FSO and will work in conjunction with responsible persons to monitor the safety of common parts of relevant buildings. 2.3.1.2.2 Housing Act 2004 (the Act) The Act imposes a duty on local authorities to keep the housing conditions in their area under review. Local authorities may inspect the common parts of residential buildings, where they consider it appropriate to do so. In reviewing the state of the building, they must consider the 29 hazards prescribed by the Act, which include fire, noise, and structural collapse. The local authority is required to take enforcement action where a Category 1 hazard (most serious) is identified and may do so at its discretion in respect of a Category 2 hazard. Such enforcement action can include improvement notices, emergency prohibition orders (preventing specified uses of the property), and demolition orders.

30

Fire Safety Design for Tall Buildings

Under this Act, the Local Government Group publishes fire safety guidance to assist local authorities in understanding their responsibilities. 2.3.1.3 BS 7974:2019 BSI (2019) BS7974: 2019, “Code of Practice on Application of Fire Safety Engineering Principles to the Design of Buildings,” is a British Standard providing a framework for designs that protect people, property, and the environment from fire. It contains guidance and information on how to undertake quantitative and detailed analyses of specific aspects of the design. They don’t preclude the use of appropriate methods and data from other sources. It includes below parts: Part 1: Initiation and development of fire within the enclosure of origin Part 2: Spread of smoke and toxic gases within and beyond the enclosure of origin Part 3: Structural response and fire spread beyond the enclosure of origin Part 4: Detection of fire and activation of fire protection systems Part 5: Fire service intervention Part 6: Evacuation Part 7: Probabilistic fire risk assessment The primary users will be fire safety engineering practitioners. Other users include members of the fire and rescue service, structural engineers, architects, government departments, universities (for teaching and research), regulators, and related industries such as insurers and systems engineers. 2.3.1.3.1 PD 7974-3:2019 BSI PD7974-3 (2019), “Application of fire safety engineering principles to the design of buildings—Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3),” is primarily used for Structural Engineers in structural fire design. PD 7974-3 (2019) considers the following issues: • The conditions within a fire enclosure and their potential to cause the fire to spread by way of recognized mechanisms and routes. • The thermal and mechanical responses of the enclosure boundaries and its structure to the fire conditions. • The impact of these anticipated thermal and mechanical responses on adjacent enclosures and spaces. • The structural responses of load-bearing elements and their effect on structural stability, load transfer, and acceptable damage.

Regulatory requirements & design principles

31

2.3.1.4 BS 9999:2017 BS 9999 (2017), “Code of practice for fire safety in the design, management and use of buildings,” gives recommendations and guidance on the design, management, and use of buildings to achieve reasonable standards intended to safeguard the lives of building occupants and fire fighters. It also provides guidance on the ongoing management of fire safety within a building throughout its life cycle, including guidance for designers to ensure that the overall design of a building assists and enhances the management of fire safety. It is applicable to the design of new buildings and also to alterations, extensions, and changes of use of an existing building. It is not applicable to individual dwelling houses and might have only limited applicability to certain specialist buildings and areas of buildings (e.g. hospitals and areas of lawful detention). The primary users will be architects, fire safety engineers, fire risk assessors, building control, installers of fire and smoke alarms, sprinklers, and smoke and heat control systems. 2.3.1.5 BS 5950-8:2003 BS5950-8, British Standards Institution (2003), “The structural use of steelwork in buildings, Part 8: code of practice for fire-resistant design,” gives recommendations for evaluating the fire resistance of steel structures. Methods are given for determining the thermal response of the structure and evaluating the protection to achieve the specified performance. 2.3.1.6 BS 476-20:1987 BS 476-20 (1987), “Incorporating Amendment No. 1. Fire tests on building materials and structures—Part 20: Method for determination of the fire resistance of elements of construction (general principles),” gives detailed specifications to determine the fire resistance of protected or unprotected elements of a building. 2.3.1.7 Design guidelines from IStructE and Steel Construction Institute In the UK, the Institution of Structural Engineers and Steel Construction Institute also issued design guidelines for fire-resistant design. They particularly deal with structural fire design issues. • IStructE (2007), Guide to the advanced fire safety engineering of structures, Institution of Structural Engineers, August 2007.

32

Fire Safety Design for Tall Buildings

• The Steel Construction Institute, Fire Resistant Design of Steel Structures, A Handbook to BS 5950: Part 8.

2.3.2 Eurocode Eurocode has probably the most detailed design guidance on fire safety design, including means of escape; fire spread; reaction to fire; the fire resistance of the structure in terms of resistance periods; the smoke and heat exhaust ventilation system; active firefighting measures such as hand extinguishers, smoke detectors, and sprinklers; and access for the fire brigade. The major design guidelines are as follows: • EN 1991-1-2 (2002), Eurocode 1. Part 1–2: General actions—actions on structures exposed to fire. • EN1992-1-2 (2004), Eurocode 2. Design of concrete structures, Parts 1–2: general rules—structural fire design. • EN 1993-1-2 (2005), Eurocode 3. Design of steel structures, Parts 1–2: general rules. Structural fire design. • BS EN 1994-1-2 (2005) Eurocode 4. Design of composite steel and concrete structures, Parts 1–2: general rules.

2.3.3 Guidelines from International Organization for Standardization 2.3.3.1 ISO 24679-1:2019(en) ISO 24679-1:2019(en), “Fire safety engineering—Performance of structures in fire—Part 1: General,” is a standard issued by International Organization for Standardization. It provides a methodology for assessing the performance of structures exposed to a real fire. It provides a performance-based methodology for engineers to assess the level of fire safety of new or existing structures. The fire safety of structures is evaluated through an engineering approach based on the quantification of the behavior of a structure for the purpose of meeting fire safety objectives and can cover the entire time history of a real fire (including the cooling phase) and its consequences related to fire safety objectives such as life safety and property protection. 2.3.3.2 ISO 16730-1 and ISO 16733-1 ISO 16730-1, “Fire safety engineering—Procedures and requirements for verification and validation of calculation methods—Part 1: General,” and ISO 16733-1, “Fire safety engineering—Selection of design fire scenarios and design fires—Part 1: Selection of design fire scenarios,” are the other

Regulatory requirements & design principles

33

two guidelines that deal with the calculation methods and selection of fire scenarios in fire safety engineering. 2.3.3.3 ISO 834-1:1999 International Organization for Standardization issued has another design code as well. ISO 834-1 (1999), “Fire-resistance tests—Elements of building construction—Part 1: General requirements,” is a similar code to BS 476-20, which deals with the guideline on fire testing.

2.3.4 US design guidance In the US, fire protection is impacted by a number of codes and standards. The most frequently used codes and standards are issued by the following professional bodies: • National Fire Protection Association (NFPA) • International Code Council • American Society for Testing and Materials (ASTM) publishes several fire protection-related standards through its E-5 committee • American Society of Civil Engineers (ASCE). There are several codes and standards published by these professional bodies. They will be briefly introduced here. 2.3.4.1 National Fire Protection Association NFPA has published several standards for fire safety design such as: • NFPA 13, Standard for the Installation of Sprinkler Systems • NFPA 72, National Fire Alarm Code • NFPA 1, Uniform Fire Code. NFPA has also published several codes and standards that cover specific aspects of fire protection and fire-related hazards. 2.3.4.2 International Code Council— International Fire Code ® (IFC ®) The International Fire Code (2018) establishes minimum regulations for fire prevention and fire protection systems using prescriptive and performance-related provisions. This code addresses extraordinary fire risks in existing buildings with retrospective requirements, but only in this limited

34

Fire Safety Design for Tall Buildings

area is there a need for alterations, as long as the building and its occupancies comply with reasonable fire-prevention provisions. 2.3.4.3 American Society for Testing and Materials (ASTM) ASTM has issued several standards related to fire safety as follows: 2.3.4.3.1 Specification for fire resistance ASTM E119-19, “Standard Test Methods for Fire Tests of Building Construction and Materials,” specifies test methods to evaluate the duration for which the types of building elements in fire. ASTM E2748-12a (2017), “Standard Guide for Fire-Resistance Experiments,” specifies methods and procedures set forth in this guide related to the conduct and reporting of fire-resistance tests obtained from particular fire-resistance specimens tested using conditions different from those addressed by Test Methods E119. 2.3.4.3.2 Specification for fire safety engineering • E1355-12 (2018), Standard Guide for Evaluating the Predictive Capability of Deterministic Fire Models • E1546-15 Standard Guide for Development of Fire-Hazard-Assessment Standards • E1591-13 Standard Guide for Obtaining Data for Fire Growth Models • E1776-16 Standard Guide for Development of Fire-Risk-Assessment Standards • E3020-16a Standard Practice for Ignition Sources. 2.3.4.4 American Society of Civil Engineers ASCE (1992), ASCE Manuals and Reports on Engineering Practice No. 78, Structural Fire Protection, Prepared by the ASCE Committee on Fire Protection Structural Division American Society of Civil Engineers, is a manual intended to provide a basis for calculation of the fire resistance of structural members. It not only focuses on design guidelines for structural fire safety design, but it also provides information on current techniques and developments to improve fire safety in buildings. It covers fire severity, response of various materials such as concrete and steel in fire, and fire protection. The manual consists of two parts: the objective of Part 1, consisting of Chapters 1–3, is to introduce the subject matter to the building design

Regulatory requirements & design principles

35

practitioner who has had no experience with fire other than in work with building codes. The material in this part is mainly descriptive. In Part 2, which consists of Chapters 4 and 5, the technical basis of the materials in Part 1 is described. This will enable those interested to obtain more knowledge about the background of the materials in Part 1. 2.3.4.5 Federal Standards and Guidelines There are also some Federal Standards and Guidelines. 2.3.4.5.1 Department of Defense (DOD) DOD UFC 3-600-01: Fire Protection Engineering for Facilities. 2.3.4.5.2 General Services Administration (GSA) • PBS-P100 Facilities Standards for the Public Buildings Service • “Fire Safety Retrofitting in Historic Buildings” by Advisory Council on Historic Preservation and General Services Administration, 1989.

2.3.5 Chinese design guidance GB50016 (2014), Code for Fire Protection Design of Buildings published by National Standard of the People’s Republic of China, is a design code for fire safety design of tall or multistorey buildings, tunnels, shopping centers, etc. It covers newly built buildings and extension or alteration to existing buildings. It stipulates the design requirements for compartmentation, evacuation, and specification of the dimension of individual member in a building to be able to comply with fire safety requirement. CECS 392 (2014), Code specification for anti-collapse design of building structures by China Association for Engineering Construction Standardization, is a design specification for the collapse prevention of buildings. It comprises a part that specifies the guidance for fire-induced collapse, and the specification requires the structure to resist fire for a sufficiently long time without collapse. Three methods are introduced: the simplified component method, the alternative load path method, and the advanced analysis for entire fire process.

2.3.6 New Zealand code NZS 3404 Part 1:1997 The New Zealand steel code (NZS, 1997) includes a section for fire safety of essential steel elements of a structure. It specifies similar formulas of Eurocodes for the maximum temperature that an element can reach before

36

Fire Safety Design for Tall Buildings

it will no longer be able to carry the design load in fire condition and therefore fail, and the time until this temperature is reached. Other design tools include formulas to estimate the variation in mechanical properties of steel with temperature.

2.3.7 Australian code AS 4100:1998 The Australian Steel Code is similar to New Zealand Steel Code except for a few minor alterations such as the use of the exposed surface area to mass ratio (ksm) instead of the section factor which is used in both New Zealand code NZS 3404. 1.5.3 and Eurocode ENV 1993-1-2.

2.4 BASIC PRINCIPLES FOR FIRE SAFETY OF TALL BUILDINGS As tall buildings become more complex with dramatic changes in building envelope and materials, it is vital to consider fire safety implications of new buildings or other construction or refurbishment projects at the concept design stage. A successful fire safety design requires an understanding of a wide range of issues and components, and the interactions between them. At all stages of the project design, the following factors need to be considered: • • • • • • • •

Regulations compliance Fire detection and suppression Fire modeling, and risk assessments Heat transfer to the structure Materials fire rating Fire protection measures—active and passive Smoke movement and smoke and heat exhaust ventilation systems People movement and means of escape.

2.4.1 Main design objective It should be bore in mind at the beginning of this book that the main objective of fire safety design is to save lives, not to prevent collapse of buildings. The primary objective is to reduce the potential for death or injury to the occupants of a building and others who may become involved, such as the fire and rescue services. It is also crucial to protect contents of the buildings and ensure that, as much as possible, the building can continue to function after the fire or that it can be repaired.

Regulatory requirements & design principles

37

Therefore, in the event of an outbreak of fire, the load-bearing elements of a building should continue to function until all occupants have escaped, or been assisted to escape, from the building. In addition, to achieve this primary objective, effective compartmentation and evocation route, as well as other factors such as means of warning, should be guaranteed in the design.

2.4.2 Main design tasks Based on the key objective, there are many aspects that need to be addressed when designing fire safety, and particularly the fire safety design for tall buildings encompasses a wide range of techniques addressing: • • • • • • •

Means of warning and escape Compartmentation and ventilation Structural fire design Smoke control, spread of smoke Active measures for fire containment and control (sprinklers) Fire safety management Human behavior in the event of a fire.

However, among them, the key design tasks for fire safety are compartmentation, evacuation, and structural fire design. These three factors affect each other; for example, when designing the evacuation route, the time of evacuation is affected by the time of failure of structural members in fire. Therefore, an effective fire protection design is also essential to evacuations. These design tasks and how to achieve them will be discussed in detail in Chapter 3.

2.4.3 Structural fire design As the structural fire design is the key for fire safety design, it is worth a brief introduction at the beginning of this book. Structural fire design determines the thermal behaviors of the structural members and finds effective means of fire protections to satisfy fire safety design objectives. First, the atmosphere temperature needs to be determined, and then, heat transfer to the structural elements must be calculated. Different levels of analysis can be used (design formula, simplified model, or sophistic finite element analysis). When the temperature field of the structure is obtained, from the combination of the mechanical and thermal loads in case of fire, the thermal behaviors the structural elements can be assessed, which allows for further assessment against a range of performance criteria in terms of deformation and structural damage.

38

Fire Safety Design for Tall Buildings

The choice of performance for design purposes will be dependent on the consequences of failure and the function of the building. For certain highprofile multistorey buildings, this may mean that no structural failure is allowed to take place during the whole duration of the fire. 2.4.3.1 Key design tasks in structural fire design For a structural fire design, there are two key design tasks: stability and integrity. 2.4.3.1.1 Stability The overall structural stability of the building as well as the primary structural elements (columns, beams, connections, load-bearing walls) should continue to function to ensure sufficient time for the evacuation of all the occupants, or until extinction of the fire. 2.4.3.1.2 Ensure compartment integrity In the event of fire, compartments should continue to be maintained integrated to limit the passage of smoke and flame. In a structural fire design, this is achieved primarily through the control of deflections of key compartment components such as slab and beam-supporting compartment wall. 2.4.3.2 Design approach There are two major design approaches that can be used for a structural fire design, namely, prescriptive-based design and performance-based design. 2.4.3.2.1 Prescriptive-based design This method is to set up safety factors by constraining design output to preestablished bounds; in other words, it is to design the structure based on fire rating of materials which is in compliance with a code-specified value. If a designer follows these rules, they will fall within the bounds, and the design can be finished. 2.4.3.2.2 Performance-based design A designer needs to first understand the level of the performance that is expected (Custer and Meacham, 1997) and then satisfy this level in the design. It includes evaluating the strength and stiffness of the structural members for a particular designed fire, and thus achieving the stability of the structure.

Regulatory requirements & design principles

39

2.4.3.3 Pros and cons of the two design methods The constraints of the prescriptive design have been noticed by most of the practicing engineers. One of the famous examples is Grenfell Tower, which simply satisfies the bounds of the fire rating if materials do not necessarily satisfy the objective of fire safety. Therefore, it can be noticed that the current design codes in most of the countries are moving toward the latter approach. Through the investigation on WTC7, NIST NCSTAR (2008) also recommends that the fire resistance of structures can be enhanced by requiring a performance objective that building fires result in burnout without partial or global (total) collapse. The prescriptive design methods for determining the fire resistance rating of structural assemblies do not explicitly specify a performance objective. The rating resulting from current test methods indicates that the assembly continued to support its superimposed load during the test exposure without collapse, however it is collapse These two design methods will be described in detail in Chapter 5.

2.4.4 Robustness of the structure in fire Although the key design objective when buildings in fire is saving lives, and not to prevent collapse, when a building is in fire, if the building collapses, for sure it will affect the safety of lives. Therefore, the collapse of the building during the fire should be prevented. In Chapter 7, this topic will be discussed in detail.

2.4.5 Fire modeling In a fire safety design, fire modeling is an important tool for effective delivery of feasible design measures. It comprises two major areas: • Modeling the atmosphere temperature induced by fire. • Modeling the thermal response of building elements (primarily loadbearing structural members or sometimes non-load-bearing members). 2.4.5.1 Modeling the atmosphere temperature induced by fire There are several ways to model the atmosphere temperature during the fire such as CFD or zone model, filed model, or simple model (by using some simplified formula to calculate the atmosphere temperature, Eurocode adopted this approach).

40

Fire Safety Design for Tall Buildings

2.4.5.2 Modeling the thermal response of load-bearing building elements 1. Simplified models If the heated temperature of the structural member is below the critical temperature, there is no failure, but if it is higher than the critical temperature, there is failure. It is a “pass or failure” criterion. The objective is then reached if the time to reach the failure is greater than the required natural fire exposure. This so-called critical temperature method is adopted in both British code and Eurocode. This method will be explained in detail in Chapter 4. 2. Finite element model Finite element modeling software (Fu, 2016) is required to model the response of the structural members in fire. The results are assessed generally in terms of deformation during the whole fire duration. The performance criteria (to measure at which level the objectives are fulfilled) in terms of deformation can be used for the assessment. Another more stringent and complicated method is the so-called multiphysics thermal–mechanical coupled modeling, which is the most accurate way to assess the behaviors of the structural member under fire. However, it is computationally expensive, and thus it is not necessary in most of the cases. 2.4.5.3 Summary As it can be seen, fire modeling is an important tool in fire safety design, and it will be discussed in detail in Chapter 6.

REFERENCES AS 4100 (1998), Building code of Australia, steel structure, standards Australian. ASCE (1992), ASCE Manuals and Reports on Engineering Practice No. 78: Structural fire protection, ASCE Committee on Fire Protection Structural Division. ASTM E119-19, Standard test methods for fire tests of building construction and materials. BBC (2017),Tehran Fire: Twenty Firemen killed as High-rise Collapses, https://www. bbc.co.uk/news/world-middle-east-38675628. British Research Establishment (1999), The Behaviour of Multi-Storey Steel Framed Buildings in Fire. British Steel Plc Swinden Technology Center. ISBN 0900206500 British Research Establishment (2004), ‘Client report: Results and observations from full-scale fire test at BRE Cardington, 16 January 2003 Client report number 215–741’, February 2004 (Accessible from: http://www.mace.manchester. ac.uk/project/research /structures/strucfire/DataBase/TestData/default1.htm) British Research Establishment (2017), Interim BRE global client report for Grenfell Tower:

Regulatory requirements & design principles

41

BS 476-20 (1987), Incorporating Amendment No. 1. Fire tests on building materials and structures, Part 20: Method for determination of the fire resistance of elements of construction (general principles). BS5950-8, British Standards Institution (2003), The structural use of steelwork in buildings, Part 8: Code of practice for fire resistant design. BSI BS 9999 (2017), Code of practice for fire safety in the design, management and use of buildings, The British Standards Institution. BSI ‘PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3)’, The British Standards Institute. BSI‘BS7974 (2019), Code of practice on application of fire safety engineering principles to the design of buildings’, The British Standards Institution. CECS 392 (2014), Code Specification for anti-collapse design of building structures by China Association for Engineering Construction Standardization. Custer, R., Meacham, B. J. (1997), Introduction to performance-based fire safety, Society of Fire Protection Engineers. EN 1991-1-2 (2002), Eurocode 1. Actions on structures, Part 1–2: General actions. Actions on structures exposed to fire. Commission of the European communities EN 1992-1-2 (2004), Eurocode 2. Design of concrete structures, Part 1–2: General rules. Structural fire design. Commission of the European communities. EN 1993-1-2 (2005), Eurocode 3. Design of steel structures, Part 1–2: General rules. Structural fire design. Commission of the European communities. EN 1994-1-2 (2005), Eurocode 4. Design of composite steel and concrete structures, Part 1–2: General rules. Structural fire design. Commission of the European communities. Engelhardt, M., Meacham, B., Kodur, V., Kirk, A. (2013), Observations from the fire and collapse of the faculty of architecture building, Delft University of Technology, DOI: 10.1061/9780784412848.101, Structures Congress 2013 Fu, F. (2016), 3D finite element analysis of the whole-building behaviour of tall building in fire. Advances in Computational Design, 1(4), pp. 329–344. Fu, F. (2017), Grenfell Tower disaster: How did the fire spread so quickly? BBC Australia. GB50016 (2014), Code for Fire Protection Design of Buildings, National Standard of the People’s Republic of China. HM Government (2019), The Building Regulations 2010-Approved Document B, Volume 1 fire safety, Dwellings’, HM Government. HM Government (2019), The Building Regulations 2010-Approved Document B, Volume 2 fire safety, Buildings other than Dwellings’, HM Government. International code council (2018), International fire code. ISBN 978-1-60983-739-6, INTERNATIONAL CODE COUNCIL, INC. Date of First Publication: August 31, 2017, U.S.A.: Publications, 4051 Flossmoor Road, Country Club Hills, IL 6047. ISO 16730-1, Fire safety engineering—Procedures and requirements for verification and validation of calculation methods, Part 1: General. ISO 16733-1, Fire safety engineering—Selection of design fire scenarios and design fires, Part 1: Selection of design fire scenarios. ISO 834-1 (1999), Fire-resistance tests—Elements of building construction, Part 1: General requirements, Edited by T. T. Lie.

42

Fire Safety Design for Tall Buildings

ISO 24679-1 (2019), Fire safety engineering—Performance of structures in fire, Part 1: General IStructE (2007, August), Guide to the Advanced Fire Safety Engineering of Structures, Institution of Structural Engineers, London. Meacham, Brian J., Park, H., Engelhardt, M., Kirk, A., Kodur, V., van Straalen, I., Maljaars, J. P., van Weeren, K., de Feijter, R., Both, K. F. (2010), Fire and collapse, Faculty of Architecture building, Delft University of Technology: Data collection and preliminary analyses. NIST (2007), Best practices for reducing the potential for progressive collapse in buildings, National Institute of Standards and Technology, Technology Administration, U.S. Department of Commerce. NIST NCSTAR (2005, December), Federal building and fire safety investigation of the World Trade Center disaster, final report of the National Construction Safety Team on the collapses of the World Trade Center Towers. NIST NCSTAR 1A (2008), Final report on the collapse of world trade center building 7, National Institute of standards and Technology, US department of commerce. NZS 3404 Parts 1 and 2:1997, Steel Structures Standard. Scoss Failure Data Sheet (2008), The Fire at the Torre Windsor Office Building, Madrid 2005. The Steel Construction Institute, Fire Resistant Design of Steel Structures, A Handbook to BS 5950: Part 8 Usmani, A. S., Chung, Y. C., Torero, J. L. (2003), How did the WTC towers collapse: A new theory. Fire Safety Journal, 38(6), pp. 501–533.

Chapter 3

Fundamentals of fire and fire safety design

3.1 INTRODUCTION In this chapter, the fundamental knowledge of fire and fire safety design will be explained. The characteristics of fire and its development are introduced at the beginning. Then, the key scenarios that affect the performance of the building members in fire—such as ventilation-controlled or fuel-controlled fire, and long-cool, short-hot fire—will be explained. In addition, the fundamentals of heat transfer, a process of the heating up of structural members due to fire, will be introduced. The basic structural fire design principles will also be explained. In fire safety design, most of the codes specify the fire resistance for building elements. The relevant information will be provided in the latter part of this chapter followed by the introduction of fire protection methods.

3.2 FIRE DEVELOPMENT PROCESS As shown in Figure 3.1, the development process of a real fire in a confined area consists of below major phases: • Ignition: ignition and smoldering of fire at very low temperatures with varying durations. • Growing phase or pre-flashover (localized fire): the duration of this phase depends mainly on the characteristics of the compartment. Still, the fire remains localized. • Flashover: this phase is generally very short. The temperature sharply increases at this stage. • Post-flashover fire (fully developed): this phase corresponds to a generalized fire for which the duration depends on the fire load and the ventilation. • Decay phase: the fire begins to decrease after burning all combustible materials completely. 43

44

Fire Safety Design for Tall Buildings

Temperature

Growth phase

Flashover

Fully developed phase

Decay phase

Ignition

Time

Figure 3.1 Temperature–time curve for fire development in a confined area.

3.3 DESIGN FIRE TEMPERATURE In structural fire design, design fire temperature curves are used to represent the atmosphere temperature (compartment temperature). As shown in Figure 3.2, there are two main fire temperature curves from current design codes that the engineers can use directly in their structural fire analysis, namely, standard fire temperature and parametric fire temperature.

Figure 3.2 Standard fire time temperature and parametric fire temperature with different opening factors (570 MJ/m2 fire load density).

Fundamentals of fire and fire safety design

45

3.1.1 Standard fire temperature–time curve As shown in Figure 3.2, standard fire temperature curve is primarily used to define the heating condition for fire test on the structural members, but the cooling phase is not included. BS476: part20 (1987) gives the formula to calculate the temperature: T = 345log10 (8t + 1) + 20

(3.1)

where T is the mean furnace temperature (in °C), t is the time (in min) up to a maximum of 360 min.

3.1.2 The parametric temperature–time curves The parametric fire is defined in Annex A of EN1991-1-2: Eurocode 1; Part 1.2 (2002) (see Figure 3.2). The parametric temperature–time curves has both heating and cooling phase. It resembles the real atmosphere temperature Its heating phase given by EN1991-1-2: Eurocode 1; Part 1.2 (2002) are obtained as follows:

(

*

*

Θ g = 20 + 1325 1 − 0.324e −0.2t − 0.201e −1.7 t − 0.472e −19t

*

)

(3.2)

where Θ g is the gas temperature in the fire compartment, t * = Γt , t = time, Γ = [O / b ]

2

[0.04 / 1160]2 ,

O = opening factor, O=

Av H w At

where At = total internal surface area of compartment [m 2] Av = area of ventilation [m 2] Hw = height of openings [m] b = thermal diffusivity, 100 ≤ b ≤ 2,000 J/m 2s1/2 K , The maximum temperature Θmax in the heating phase happens when t * = t *max:

(

t *max = t max Γ t *max = t max • Γ

)

46

Fire Safety Design for Tall Buildings

with qt ,d ⎞ ⎛ t max = ⎜ 0.2 10−3 ⎝ O ⎟⎠

(

t max = 0.2  ⋅ 10−3 • qt ,d O

)

or t lim. qt,d is the design value of the fire load density related to the total surface area At of the enclosure: qt ,d = qf ,d ×

Af ⎡ MJ/m 2 ⎦⎤ At ⎣

The following limits should be observed: 50 ≤ qt ,d ≤ 1000 ⎡⎣ MJ/m 2 ⎦⎤ t lim in case of slow fire growth rate, t lim = 25 min; in case of medium fire growth rate, t lim = 20 min; in case of fast fire growth rate, t lim = 15 min. From Equation 3.2, it can be seen that the opening factor plays an important role in the atmosphere temperature. As it can be seen from Figure 3.2, under the same fire load density of 570 MJ/m 2 , when the opening factor increases from 25% to 100%, the parametric fire temperature curves change significantly, the fire changes from “long-cool” to “short-hot.” These two fire scenarios have distinctive effects on the response of structural members. This will be explained in detail Section 3.4.

3.1.3 Summary From the above, it can be seen that, compared to standard fire curve, parametric fire curve has a cooling phase. The parametric fire curve more closely represents the real fire temperature than standard fire curve. For designing the fire resistance capacity of each individual member, the standard fire curve shall be used, as it gives more conservative results. For determining the behavior of a structural member, especially behavior of the wholebuilding or a frame, the parametric fire curve shall be used, as it gives more realistic fire temperature development in a compartment. 3.4 DESIGN FIRE IN A COMPARTMENT Fire conditions depend on many factors such as the building’s function (offices, car parks, etc.) and the materials (such as concrete or steel) used.

Fundamentals of fire and fire safety design

47

The complexity of tall buildings can cause different fire scenarios. For instance, depending on the compartmentation and structural layout of buildings, different fire temperatures can be reached. Different factors (such as fuel or ventilation) can also affect the development of the fire. The duration of the fire is determined by the opening factors and the type of fuels (such as the furniture and decorations). As indicated in PD7974-part 3 (2019), the design fires in a compartment need to consider below key factors: • • • • •

Nature of the fire load. Fire load density. Thermal properties of the enclosure. Extent of ventilation. Ceiling height.

For determining fire conditions in each compartment, in PD7974-part 3(2019), three methods are given: 1. Standardized models with recognition of fire conditions and occupancy type. 2. Experimental data appropriate to the compartment and occupancy type. This method primarily uses the means of existing experimental tests data to determine the fire conditions. However, although it relies on a vast amount of test data, the required data may not be available for certain specific cases most of the time. 3. Engineering calculations based upon experimentally calibrated methods. This method primarily uses experimentally calibrated design rules to determine the fire conditions. It is simpler and more efficient for an engineer. It will be described in the following sections.

3.4.1 Characterization of compartment For determining the design fire, the compartment needs to be characterized first. It comprises two major characterizations: enclosure and opening of the compartment. 3.4.1.1 Characterization of fire enclosure In compartment design, compartment walls at the boundary of the compartment are purposely designed to prevent the spreading of fire. The compartment should be evaluated on fire conditions and the potential for fire spread. The fire resistance of compartment walls can be tested in a standard furnace.

48

Fire Safety Design for Tall Buildings

In addition, in a building, all solid boundaries can slow down the spread of fire to some extent. The boundary members of a compartment are particularly important in terms of growth rate of the fire and the development of temperatures within the space. Therefore, fire conditions correlate with the capacity of initial enclosing surfaces to remain imperforate to the spread of fire. For characterizing the enclosure of fire origin, PD7974 Part 3 (2019) offers the following guidance: • When predicting pre-flashover fire conditions, the horizontal and vertical surfaces immediately surrounding the fire should be considered as the enclosure. • The enclosure can also contain openings which, although immediately and directly open to the passage of fire and heat, may be characterized as part of the enclosing boundaries. • After flashover has occurred, solid boundaries may be assumed to remain imperforate to fire as long as there are no openings created on their surfaces due to the mechanical force of the fire. • As the definition of the enclosing surfaces is changed by the creation of openings, the fire conditions may need to be re-examined and fire spread routes re-evaluated. 3.4.1.2 Characterization of openings Openings such as door, windows, and vents play an important role in fire development in a compartment, as they permit airflow to the fire and ventilation of heat. The characterization of openings includes their size, shape, and extent. Where a combination of fixed opening conditions is possible for the fire enclosure (e.g. some doors are open, and some doors are closed), options that are most conducive to fire spread should be considered. The assumption that all openings are initially open may not necessarily be the worst case. The following guidance is offered by PD7974 Part 3 (2019): • Doors should be assumed open if the enclosure has no other openings. • Doors should be assumed closed if the enclosure has other openings. • All enclosure surfaces (including glazed openings) may be assumed to be imperforate for the duration of the fire. 3.4.1.3 Duration of fire to be adopted in design The appropriate duration of the fire is influenced by the following parameters: • Occupancy; • Presence of automatic sprinkler system;

Fundamentals of fire and fire safety design

49

• Height of enclosure above ground level; • Depth of enclosure below ground level.

3.4.2 Fuel-controlled and ventilation-controlled fire In the fire development process, the amount of fuels (combustible material in a compartment) and the opening factor are the two key influential factors. As the amount of fuels dose not vary too much for certain types of buildings (either commercial or residential), the percentage of openings is the main variable for determining worst-case conditions of the compartment. As it has been explained, flashover is the transition from the localized fire to the compartment room fire. However, for rooms with very large window openings, too much heat may be released through the windows, which leads to the occurrence of flashover. Even at the post-flashover stage, the rate of the combustion also depends on the size and shape of the ventilation openings. A ventilation-controlled fire occurs when there is not enough air to support the complete combustion process of the fuel in a compartment. Therefore, the fire will extinct when the available oxygen runs out. A fuel-controlled fire is where there is adequate amount of air but not enough fuel to support the combustion process of the fuel in the fire compartment. Therefore, the fire will extinct when all the fuels are burnt. As introduced in Chapter 2, Test 4 of Cardington tests is to simulate the fuelcontrolled fire with all windows closed. As can be seen in Equation 3.2, the opening factor O included in the equation is essential to determine whether a fire is fuel controlled or ventilation controlled. Large opening factor allows more oxygen to enter into the compartment, thus leading to fuel-controlled fire. Smaller opening factor allows less oxygen to enter into the compartment; therefore, ventilationcontrolled fire is likely to happen.

3.4.3 Long-cool and short-hot fire Depending on the opening factors, when exposed to fires, the structure responds in two distinct ways: Under “short-hot” fire, the unprotected steel reaches the temperature similar to the fire temperature in atmosphere. However, for a concrete slab, due to its lower thermal conductivity, under short fire exposure, it reaches an average temperature only marginally higher than ambient conditions. So, steel and concrete material behave distinctively in “short-hot” fire. In addition, large deflections develop in a very short time. Research by Lamont et al. (2004) shows that large deflections develop in a very short time, which may result in early compartmentation failure.

50

Fire Safety Design for Tall Buildings

Under “long-cool” fire, structural members experience much longer duration of heating, and so both steel and concrete members can reach considerably high average temperatures. For steel composite structural members, in “short-hot” fire, the composite floor structure will experience higher temperature gradients across the cross-section. Therefore, thermal bowing is greater than that in the “longcool” fire. The long duration of the “long-cool” fire results in higher temperatures in the concrete and the steel. Due to higher temperatures achieved in concrete slabs, there is much less tension in the slab with growing compression toward the end of heating. Therefore, it can be concluded that in most of the cases, the worst-case fire scenario in terms of the structural response is often “short-hot” fire.

3.4.4 Fully developed fire Fully developed fire is the stage after flashover where all combustibles are fully burnt. For determining the atmosphere temperature for a fully developed fire, using parametric fire temperature from Eurocode is the simplest method. Zone model and CFD model can also provide more accurate results.

3.4.5 Localized fire In large open-space buildings such as airport terminals, railway stations, large industrial halls, car park, and some commercial buildings, post-flashover fire is unlikely to occur. Here, fuel-controlled fire would be more likely. The common scenario in this type of plan layout is that the fire starts as a localized fire. As shown in Figure 3.3, a localized fire is a pre-flashover fire which is expected to burn locally, and the temperature of a structural member is not uniform along the plume.

HOT LAYER Mass Flow Out

Layer interface COLD LAYER Smoke Layer Height

Figure 3.3 Schematic drawing for localized fires.

Fundamentals of fire and fire safety design

51

Eurocode assumes that the heating is uniform which in certain circumstance is not conservative, especially in the case of localized fires. High-intensity fire exposure effects may cause more severe damages to the structural members. The thermal response of such fires depends not only on the flame temperature but also on the magnitude and dimensions of the flames. Zhang et al. (2003) suggest that the failure mode of a beam may be different if it is exposed to a localized fire instead of the standard fire curve. Localized fire could still expose structures to severe thermal conditions, even though the mean temperature in the enclosure is low. It can significantly weaken the load-bearing capacity of a structural member in comparison to a uniform fire exposure. For determining the atmosphere temperature for a localized fire, using a simplified plume model Eurocode 1 (EN 1991-1-2, 2002) is a common method, which will be introduced here. Other more accurate methods would be using one-zone model and CFD model, which will be introduced in Chapter 6. 3.4.5.1 Calculation of thermal action of a localized fire from Eurocode EN 1991-1-2 (2002) specifies the thermal action of a localized fire with the consideration of the flame length. The flame lengths Lf of a localized fire (see Figure C.1) is given by 2

Lf = −1.02D + 0.0148Q 5

(3.3)

• When the flame does not impact the ceiling of a compartment (Lf < H; see Figure C.1) or in case of fire in open air, the temperature Θ(z) in the plume along the symmetrical vertical flame axis is given by Θ ( z ) = 20 + 0.25 (Qc )



5 3

< 900

(3.4)

where D is the diameter of the fire [m], see Figure 3.4, Q is the rate of heat release [W] of the fire, Qc is the convective part of the rate of heat release [W], with Qc = 0.8 Q by default, z is the height [m] along the flame axis, see Figure 3.4, H is the distance [m] between the fire source and the ceiling, see Figure 3.4. • When the flame impacts the ceiling (Lf > H; see Figure 3.5), the heat flux h_ [W/m 2] received by the fire-exposed unit surface area at the level of the ceiling is given by h_ = 100,000 if y ≤ 0.30 h_ = 136,300–121,000 y if 0.30 < y < 1,0 (C.4) h_ = 15,000 y−3,7 if y > 1.0

52

Fire Safety Design for Tall Buildings

Flame axis

H Lf z

D

Figure 3.4 Localized fire (flame is not impacting the ceiling). (Figure C.1 of EN 1991-1-2 (2002), Eurocode 1. Actions on structures, Part 1-2: General actions. Actions on structures exposed to fire. Commission of the European communities 2002; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].) Flame axis

Lh r

H

D

Figure 3.5 Localized fire (flame is impacting the ceiling). (Figure C.2 of EN 1991-1-2 (2002), Eurocode 1. Actions on structures, Part 1-2: General actions. Actions on structures exposed to fire. Commission of the European communities 2002; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].)

Fundamentals of fire and fire safety design

53

3.4.6 Traveling fire As introduced in the proceeding sections, in current fire safety design, homogeneous compartment temperature assumption is made. However, fire in large compartments such as tall buildings features in non-uniformity. The fire that burns locally may spread across entire floor plates over a period of time (Rein, 2007). This kind of fire scenario is called traveling fire. The experimental study of Horová (2013) shows that during traveling fire, a high degree of non-uniformity is observed in temperature. The maximum degree of heterogeneity is reached with a temperature difference up to 400°C. A temperature difference of up to 300°C can also be observed between the center and the corners of a fiber-reinforced concrete slab. As the fire spreads in the compartment, gas temperature under the ceiling fluctuates significantly, which affects structural member. In real life, traveling fires have been observed in several fire incidents, for example, the World Trade Center Towers (NIST, 2005), and the Windsor Tower (Fletcher et al., 2006) in Madrid in 2005. As explained by Ellobody and Bailey (2011), the peak temperature changes as the fire dynamic fields with cyclic heating and cooling can appear. The pattern of cyclic temperature changes can cause cyclic deflection, leading to a dangerous scenario.

3.4.7 Fire scenarios for tall buildings As it has been discussed in the proceeding sections, localized and traveling fires dominate fire scenarios of tall buildings. The WTC7 fire (NIST, 2008) showed that fires in open floorplan offices travel through large compartments generating both areas of intense localized heating and of slow pre-heating, as well as areas of cooling. These occur simultaneously within the floor, thus producing both long-cool fires and short-hot fires (Lamont et al., 2004) as well as asymmetries with differential thermal expansion. The effects of these heterogeneities should be taken into consideration when translating temperature into heat fluxes to define the thermal loading in tall building design. Therefore, when designing a tall building in fire, the whole building behavior needs to be analyzed to understand truly how a system will perform under fire loading.

3.5 FIRE SEVERITY Fire severity is a way to determine the destructive impact of a fire, the temperatures that could cause failure of the structure. From a series of compartment tests, Ingberg (1928) suggested that fire severity of a real fire could be calculated by considering equivalence of the areas under the standard fire temperature curve and the real fire temperature curve in an compartment

54

Fire Safety Design for Tall Buildings

7HPSHUDWXUH

above a base of either 150°C or 300°C as shown in Figure 3.6. It allows engineers to assess the severity of a fire in compartment based on the standard fire time–temperature curve. This method takes several factors into consideration such as ventilation, fire loads, and compartment size. It is a correlation between fire load measured in tests as load per unit floor area and the standard fire resistance periods. Therefore, fire severity is most often defined in terms of the period of exposure to the standard test fire. The design of fire severity can be determined based on a complete burnout fire or the equivalent time of a complete burnout fire. The equivalent fire severity is the time of exposure to the standard fire that would result in the load-bearing capacity being the minimum occurring in a complete burnout of the fire cell. Law (1971) used a time equivalent method to define the equivalent fire severity using the time of exposure to the standard fire, which would result in the same maximum temperature in a protected steel member as would occur in a complete burnout of the fire test, as shown in Figure 3.7.

6WDQGDUGrUH

 5HDOrUH 

Equivalent time te,d

7LPH

Figure 3.6 Method of Ingberg (1928) to determine fire severity. 7HPSHUDWXUH

*DV WHPSHUDWXUH

6WDQGDUGrUH 0HPEHU WHPSHUDWXUH

5HDOrUH

7LPH Equivalent time t/e,d

Figure 3.7 Equivalent time method to determine fire severity Law (1971).

Fundamentals of fire and fire safety design

55

3.6 FIRE LOAD Fire loading is a measurement to determine the potential severity of a fire in a given space. It describes the amount of combustible material in a building or confined space and the amount of heat this can generate. The more the flammable materials in a space, the higher the fire load and therefore the faster the fire spreads. It is the heat output per unit floor area, often in kJ/m 2 , calculated from the calorific value of the materials present. A room with no furnishings and concrete walls will have a fire loading of nearly zero. If the fire spreads into such a room from elsewhere, it will find nothing to feed on. However, the presence of anything that is flammable such as furniture, electrical appliances or computer equipment, wood panels, acoustic tile, carpeting, curtains, or wall decorations will increase the fire loading. Buildings under construction or renovation tend to carry high fire loads in the form of construction materials, solvents, and fuel for generators.

3.6.1 Fire load calculation from Eurocode 1 Eurocode 1, EN 1991-1-2 (2002) provides the formula to calculate the fire load. Design value may be determined either from a national fire load classification of occupancies or by performing a fire load survey. The design value of the fire load qf,d is defined as follows: qf ,d = qf ,k × m × δ q1 × δ q 2 × δ n ⎡⎣ MJ/m 2 ⎤⎦

(3.5)

where m is the combustion factor, δq1 is a factor taking into account the fire activation risk due to the size of the compartment (see Table E.1 of Eurocode 1, EN 1991-1-2, 2002), δq2 is a factor taking into account the fire activation risk due to the type of occupancy (see Table E.1 of Eurocode 1, EN 1991-1-2, 2002), δn is a factor taking into account the different active firefighting measures (sprinkler, detection, automatic alarm transmission, firemen), δf,k is the characteristic fire load density per unit floor area [MJ/m²].

3.6.2 Fire load density from Eurocode 1 In determination of fire load, factors such occupancy and floor area are used to characterize fire load densities qf ,k ⎡⎣ MJ/m 2 ⎦⎤ , as given in Table 3.1, which shows fire load densities qf ,k ⎡⎣ MJ/m 2 ⎤⎦ for different occupancies.

56

Fire Safety Design for Tall Buildings

2 Table 3.1 Fire load densities qf ,k ⎡⎣ MJ/m ⎤⎦ for different occupancies

Occupancy

Average

80% Fractile

Dwelling Hospital (room) Hotel (room) Library Office Classroom of a school Shopping center Theater (cinema) Transport (public space)

780 230 310 1,500 420 285 600 300 100

948 280 377 1,824 511 347 730 365 122

Note: Gumbel distribution is assumed for the 80% fractile. Source: Reproduced from Table E.4 of EN 1991-1-2 (2002), Eurocode 1. Actions on structures, Part 1-2: General actions. Actions on structures exposed to fire. Commission of the European communities 2002; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/ Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].

3.7 FIRE SPREAD In tall buildings, the fire spreads in different ways. The direction and the speed of its spread are determined primarily by the compartment of the buildings. The potential for the spread of fire from an enclosure will be influenced by the thermal and mechanical responses of the enclosure’s boundaries (walls, roof, doors, windows). The thermal and mechanical responses of boundary elements may be evaluated individually, subject to checking for interaction effects between adjacent structural members. For example, the thermal bowing of walls may affect the support or loading capacity of the enclosure’s roof.

3.8 ROUTES OF FIRE SPREAD Once the compartment is characterized, the designer should identify all the possible routes of fire transmission through the boundary surfaces. Figure 3.8 (PD 7974-3, 2019) illustrates some of the most common routes of potential fire spread. These routes of fire spread should be examined as a series of direct spread mechanisms. Ideally, all the potential routes of fire spread from the enclosure should be investigated and the minimum time for fire spread determined. However, design effort may be reduced in situations where expert judgment can identify the routes that are susceptible to the most rapid fire spread. It should be remembered that the determination as

Fundamentals of fire and fire safety design

57

to whether or not fire spread takes place will be influenced by conditions both within the fire enclosure and within the adjacent enclosures.

3.8.1 Horizontal spread of fire As shown in Figure 3.8, horizontal spread of fire is primarily through wall (if it has relatively low fire resistance), opening in the wall, ceiling, and void or duct in ceiling or floor.

3.8.2 Vertical spread of fire As shown in Figure 3.7, vertical spread of fire can be primarily divided into two categories: internal and external.

Through wall or openings created in wall

Spread mechanism : conduction, convection, direct pyrolysis (collapse)

Through fixed openings

Spread mechanism : convection, radiation direct pyrolysis, mass transfer

Along or through vertical duct

Through floor

Spread mechanism : conduction, convection, direct pyrolysis (collapse)

Along or through horizontal duct

Spread mechanism : conduction, convection, direct

Within roof

Figure 3.8 Routes of fire transmission. (Reproduced from Figure 4 of BSI “PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3),” The British Standards Institute; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].) (Continued )

58

Fire Safety Design for Tall Buildings

Below floor

Above ceiling

Spread route: 1.Enclosure to ceiling void 2.Void to adjacent enclosure

Spread route: 1.Enclosure to floor void 2.Void to adjacent enclosure

Over external roof

Spread route: 1.Enclosure to roof 2.Roof to adjacent enclosure

Spread route: 1.Enclosure to outside 2.Outside to adjacent enclosure

Via windows

Spread route: 1.Enclosure to facade 2.Facade to adjacent enclosure

Spread route: 1.Enclosure to facade 2.Facade to adjacent enclosure Within facade or building envelope

Surface of facade

Flying brands

Exposed building

Spread route: 1.Enclosure to roof outside 2.Roof flames through external envelope,e.g. window

Radiation

Spread route:radiation/mass transfer/direct pyrolysis

Via external route

Figure 3.8 (Continued) Routes of fire transmission. (Reproduced from Figure 4 of BSI “PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3),” The British Standards Institute; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].)

Fundamentals of fire and fire safety design

59

3.8.2.1 Fire spread through ducts, shafts, and penetrations (internal) For the 19-floor Windsor Tower in Spain (Fletcher, 2006), the fire spread to most parts of the building within 7 h. The fire spread internally vertically through ducts, shafts, penetrations, etc. A fire of this nature will generally propagate extremely quickly without any hope of being controlled by sprinklers and has the potential of almost simultaneously compromising the life of everyone remaining within the building. 3.8.2.2 Fire spread through façade As introduced in Fu (2017), in Grenfell Tower fire, the fire spread primarily from the façade. The fire originated inside a room can also spread through the window to the façade and subsequently spread to other floors. The fire can spread either through the exterior of the façade or the interior gaps. To avoid the first route of spread, the material used in the façade should have sufficient fire resistance. As shown in Chapter 2, if the façade is flammable, it will accelerate the spread of fire. To avoid the second route of spread, sufficient fire blocks should be designed to stop the spread of fire through gaps. Apart from Grenfell Tower, vertical spread of fire through the façade has also been noticed in other fire incidents. In CCTV tower fire in China, the fire spread to most parts of the building in around 15 min. It spread predominantly through cladding following an ignition in the cladding from a firework.

3.9 STRUCTURAL FIRE DESIGN As mentioned in Chapter 2, structural fire design is one of the key fire safety design tasks. The basic principles will be introduced in this section, and more specific design methods will be introduced in detail in Chapter 4. The objective of structural fire design is to determine the thermal behavior of the structural members and finding effective means for fire protections to satisfy fire safety design objectives. Structural fire design generally consists of the following: • Assessment of thermal response and structural response for different types of building members; • Design structural system and its various components including supports and joints; • Choice of size of structural members and fire protections with specified thermal and mechanical properties.

60

Fire Safety Design for Tall Buildings

Characterise enclosure and its openings (clause 5)

Quantify fire conditions temperature / heat flux, pressure (clause 9)

Evaluate thermal response of element under consideration (clause 10)

Evaluate mechanical response (including collapse) of element under consideration (clause 11, 12, and 13)

Conclude structural response

Conclude time to fire spread

Figure 3.9 (Figure 12 of BSI “PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3),” The British Standards Institute.) (Permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected])

PD7974-3 (2019) specifies the procedure for designing structural members to avoid fire spread. The first step “characterize the enclosure and its opening” has been already introduced in Section 3.4.1. The remaining steps will be introduced in the following sections (Figure 3.9).

3.9.1 Determine the compartment temperature (design fire) The first task in structural fire design is to determine design fire based on fire scenarios. Normally the design fire should be applied in one fire compartment of the building at a time, unless otherwise specified in the design fire scenario. As introduced in Section 3.3, the simplest way to predict the time– temperature curve for a fire compartment is a standard fire temperature model in Eurocode. Parametric fire curve is closer to a real fire in predicting the heating rate and the maximum temperature of the atmosphere

Fundamentals of fire and fire safety design

61

inside the fire compartment. Engineers need to assess the fire load (the quantity and type of combustible material), the ventilation, and the thermal characteristics of the compartment linings. These variables can be used to calculate the fire temperature per Eurocode. Using engineering software such as Zone model and computational fluid dynamics is a more accurate way.

3.9.2 Determine the thermal response of structural members During the process of fire, heat will be transferred to the building elements through the so-called heat transfer process. The thermal response of each structural member can be worked out based on the basic thermodynamics’ principles. Heat transfer is comprised of three processes: conduction is the mechanism of heat transfer in solid materials, in the steady-state situation; convection is the heat transfer by the movement of fluids, either gases or liquids; radiation is the transfer of energy by electromagnetic waves. The increase in temperature for both internal unprotected or protected steelwork and concrete structural elements can be obtained from the fire tests given in BS 476: Parts 20–22 (1987). Eurocode 3 (2005) and Eurocode 4 (2005) give the formulas to determine the increase in temperature in both protected and unprotected structural members. These formulas are based on the principles of heat transfer which will be briefly introduced here.

3.9.3 Heat transfer 3.9.3.1 Thermodynamics of heat transfer Thermodynamics is the study of macroscopic continuum. Following energy conservation rules, a system of fixed mass must remain at a constant total energy if it is isolated from its surroundings. Under fire, only internal energy (U) will be significantly changed (excluding kinetic, electrical, etc.). From the first law of thermodynamics, under fire we can get: U2 − U1 = Q + W

(3.6)

where U is the internal energy, Q is the heat, W is the work. The rate of change in the internal energy will be dU  +W  =Q dt

(3.7)

62

Fire Safety Design for Tall Buildings

In heat transfer, heat can flow in two distinct mechanisms due to the temperature difference as per the Fourier law: 1. Conduction equation q = − λ∇θ

(3.8)

where q is the vector of heat flux per unit area, λ is the thermal conductivity, θ is the temperature. The conservation of energy with the Fourier law requires ⎛ ∂θ ⎞ Pc ⎜ ⎟ =   −∇ ( λ∇θ ) + Q ⎝ ∂t ⎠

(3.9)

where ρ is the density, c is the specific heat, t is the time, Q is the internal heat generation rate per unit volume. Appropriate boundary conditions and initial conditions are needed to solve this equation. 2. Convection and radiation Heat transfer to the surface of the structural member in a fire involves both convection and radiation. The net heat flux to the surface of the structural member can be expressed as follows: Qnet = Qnet,c + Qnet,r

(3.10)

where Qnet,c is the net convective heat flux per unit surface hnet , c = α c (θ g − θ m ) ,  net,r is the net radiative heat flux per unit surface hnet,r = _ ε m ε f σ Q ⎡ ⎤ ⎣(θ g + 273)4 − (θ m + 273)4 ⎦ , where αc is the convection heat flux coefficient, θg is the gas temperature, θm is the surface temperature of the structural member, m is the surface emissivity of the structural member, εf is the emissivity of the fire, σ = 5.67 × 10−8 W/m 2 K4 is the Stefan–Boltzmann constant.

Fundamentals of fire and fire safety design

63

3.9.3.2 Eurocode formula to determine member temperature 3.9.3.2.1 Unprotected steel Section For unprotected steel section, the increase in temperature within a small time interval is given by BS EN 1993-1-2: Eurocode 3 (BSI, 2005) and BS EN 1994-1-2: Eurocode 4 (BSI, 2005) as follows: Δθ a,t = ksh

Am / V  hnet Δt ca ρ a

(3.11)

where Δθ a,t is the increase of temperature, Am/V is the section factor for unprotected steel member, Am is the exposed surface area of the member per unit length, V is the volume of the member per unit length, ca is the specific heat of steel, ρ a is density of the steel, hnet is the designed value of the net heat flux per unit area, Δt is the time interval, ksh is the correction factor for the shadow effect. 3.9.3.2.2.1 PROTECTED STEEL SECTION

For protected steel section, the increase in temperature within a small time interval is given by BS EN 1993-1-2: Eurocode 3 (BSI, 2005) and BS EN 1994-1-2: Eurocode 4 (BSI, 2005) as follows:

Δθ a,t

⎫ ⎧ λp ⎪ ⎪ d p Ap ⎛ 1 ⎞ =⎨ ⎜ ⎟ θ − θ a,t ) Δt ⎬ − exp(Φ 10) − 1 Δθ g ,t (3.12) Φ ( g ,t ⎪ ⎪ ca ρ a V ⎝⎜ 1 + 3 ⎠⎟ ⎩ ⎭

{

where Φ=

cp ρ p d p Ap / V ca ρ a

θ a,t is the temperature of the steel at time t, Δθ a,t is the increase in temperature, Δθ g ,t is the gas temperature at time t, Δθ g ,t is the increase in the gas temperature, Ap/V is the section factor for protected steel member, ca is the specific heat of steel,

}

64

Fire Safety Design for Tall Buildings

1 Hp 2 254mm × 102mm × 28kg/m UB a) Exposed stanchion Hp/A=250m–1

1 Concrete slab 2 Hp 3 254mm × 102mm × 28kg/m UB b) Partially exposed beam Hp/A=220m–1

Figure 3.10 Calculation of section factor. (Reproduced from Figure 12 of BSI “PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3),” The British Standards Institute; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].)

cp is the specific heat of fire protection material, ρ a is the density of the steel, ρ p is the density of the fire protection material, d p is the thickness of the fire protection material, λ p is the thermal conductivity of the fire protection material, Δt is the time interval. As shown in Figure 3.11, under the same standard fire temperature, there is slight temperature difference between protected and unprotected structural steel members.

Fundamentals of fire and fire safety design

65

3.9.3.2.1.2 SECTION FACTOR

As shown in Figure 3.10, the section factor is one of the important factors in calculating the temperature of the steel members. It is defined based on both the geometry and configuration of the members used in a building. For unprotected members, the heated perimeter A m depends on the type of insulation. For protected members, the heated perimeter Am depends on the type of insulation, for example, sprayed insulation or intumescent paint which is applied to the section profile or board insulation which boxes the section and on the number of sides of the members exposed to the effect of the fire.

3.9.4 Material degradation at elevated temperatures The strength for the Material properties of steel and concrete start to lose at elevated temperatures. Eurocode gives the material degradation for both concrete and steel material, which will be introduced here. 3.9.4.1 Degradation of steel material in fire Figure 3.12 shows the stress–strain curves of steel at different temperatures from EN1994-1-2 Eurocode 4 (2005). The loss of strength can be illustrated by the amount of the stress that the member is able to withstand before reaching 2% strain.

Figure 3.11 Temperatures of steel members (protected and unprotected members).

66

Fire Safety Design for Tall Buildings

Figure 3.12 Stress–strain relationships of structural steel at elevated temperatures, strain-hardening included.(Figure A.1 of EN1994-1-2 (2005), incorporating corrigendum July 2008, Eurocode 4 Design of composite steel and concrete structures, Part 1-2: General rules. Structural fire design; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: cservices@ bsigroup.com.)

3.9.4.2 Degradation of concrete material in fire Figure 3.13 is the stress–strain relationship of concrete from EN1994-1-2, Eurocode 4 (2005). The temperature shows a significant effect on the stress–strain relationships of concrete.

3.9.5 Design values of material properties under fire In structural fire analysis, an Engineer should consider the material degradation mentioned in the proceeding sections. The design values of materials’ mechanical properties in fire Xd,fi are defined in Formula 2.1 of Eurocode 2 (EN1992-1-2, 2004) as follows: Xd , fi =

kθ Xk γ M , fi

(3.13)

Fundamentals of fire and fire safety design

67

Figure 3.13 Behavior of compressive concrete material. (Figure B.1 of EN1994-1-2 (2005), incorporating corrigendum July 2008, Eurocode 4 Design of composite steel and concrete structures, Part 1-2: General rules. Structural fire design; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].)

where Xk is the characteristic value of a strength or deformation property for normal temperature design to EN 1992-1-1. k θ is the reduction factor for a strength or deformation property (Xk,θ / Xk) dependent on the material temperature (see Figure 3.14). It shows the reduction factor for structural steel members defined by EN1994-1-2 Eurocode 4 (2005). γM,fi is the partial safety factor for the relevant material property in the fire situation.

3.9.6 Design of structural members in fire 3.9.6.1 Mechanical design approaches of structural members in fire Eurocode 1 (EN 1991-1-2, 2002) specifies the requirements for mechanical design of structural members in fire. It should start with a temperature analysis; when performing the temperature analysis of a member, the

68

Fire Safety Design for Tall Buildings

Figure 3.14 Reduction factors for stress–strain relationships allowing for strain-hardening of structural steel at elevated temperatures. (Figure A.2 of EN1994-1-2 (2005), incorporating corrigendum July 2008, Eurocode 4 Design of composite steel and concrete structures, Part 1-2: General rules. Structural fire design; permission to reproduce and extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: [email protected].)

position of the design fire in relation to the member should be taken into account. Either standard fire or parametric fire temperature can be used in the analysis. The difference between them is whether the temperature analysis of the structural members is made for the full duration (including the cooling phase) or not. After the temperature analysis, the mechanical analysis should be performed for the same duration as in the temperature analysis. The analysis can be performed in three domains. 1. In the time domain: t fi,d > t fi,requ 2. In the strength domain: Rfi,d ,t > Efi,d ,t 3. In the temperature domain: Θd ≤ Θcr,d

Fundamentals of fire and fire safety design

69

where t fi,d is the design value of the fire resistance, t fi,requ is the required fire resistance time, R fi,d,t is the design value of the resistance of the member in the fire situation at time t, Efi,d,t is the design value of the relevant effects of actions in the fire situation at time t, Θd is the design value of material temperature, Θcr,d is the design value of the critical material temperature. The specific mechanical design approaches in both temperature domain and strength domain will be introduced in Chapter 4. 3.9.6.2 The acceptance criteria in designing structural members for tall buildings For designing the structural element of a tall building under fire, an engineer can use several acceptance criteria as follows: • The overall stability of the structure is maintained. This indicates no local collapse or global collapse will be triggered due to the failure or deflection of the structural member to be designed. • Deflections of the slab or beams on any lines of compartmentation are within the allowance of the deflection of the compartment wall where relevant. This is for the integrity of the compartment to ensure that the compartment wall is still functioning during the fire. • The maximum tensile plastic strain at the level of the reinforcement is less than 5%. This is conservative in comparison with normal elongation values of 13%–15% for steel reinforcement but allows for the smeared cracking effect. • In addition, the relative deflections of the main beams will be checked to ensure that deflections do not exceed span/20 at the relevant fire resistance period required for that area of the structure. A similar check for the slab should also be made. The acceptance criteria can be achieved through a detailed structural fire analysis. Some criteria may be considered satisfied where the minimum thickness of walls or slabs is specified in accordance with Table 5.3 of EN 1991-1-2 (2002). 3.10 FIRE RESISTANCE As stated in BS 476-20 (1987), “Fire resistance is the time which an element of building construction is able to withstand exposure to a standard temperature and pressure regime without a loss of its fire separating function

70

Fire Safety Design for Tall Buildings

or load bearing function or both.” Fire resistance is a measure of the ability of the structure to resist collapse and to prevent the spread of fire during exposure to a fire of specified severity.

3.10.1 Methods to determine fire resistance The fire resistance of any individual element may be determined by a. b. c. d.

Standard fire resistance tests, Experimental large- and small-scale fire tests, Expert assessment, Quantitative analysis of fire spread mechanisms.

3.10.2 Fire resistance rating Fire resistance rating is the fire resistance assigned to a building element on the basis of a test or some other approval system. Some countries use other terms such as fire rating, fire endurance rating, or fire resistance level. These terms are usually interchangeable. Fire resistance ratings are most often assigned in whole numbers of hours or parts of hours, in order to allow easy comparison with the fire resistance requirements specified in building codes. For example, a wall that has been shown by test to have a fire resistance of 75 min will usually be assigned a fire resistance rating of 1 h.

3.10.3 Fire resistance test for load-bearing structural members Each building element needs to be assigned a fire resistance rating for comparison with the fire severity specified by codes. The most common method for assessing fire resistance is to carry out a full-scale fire resistance test. As shown in Figures 3.15 and 3.16, the building members can be placed into a fire furnace to heat up. In the meantime, it is loaded to

6WDQGDUG)LUH

Figure 3.15 PD 7974-3:2003 Standard fire resistance tests.

Fundamentals of fire and fire safety design

71

Figure 3.16 Fire test furnace. (Photo taken by the author.)

failure. The time to failure can therefore be determined. For fire resistance testing, many countries use the ISO 834-11 (2019) or have other national standards based on it.

3.10.4 Fire resistance requirements for elements of a tall building An important lesson learned from the WTC failures (NIST, 2008) and Cardington fire tests is that the prescriptive fire resistance ratings of individual building elements do not guarantee that a building system as a whole will perform adequately. A holistic performance-based fire design approach is still needed for tall buildings, which will be introduced in Chapter 5. In tall buildings, prescriptive fire resistance ratings of individual building elements do not guarantee a building system that as a whole will perform adequately. Therefore, fire resistance requirements should be based on the design fire scenarios, and the factors need to be considered are probability of fire occurrence, fire spread, fire duration, fire load, severity of fire, ventilation, compartment (type, size, geometry), and type of structural elements. In addition, the height of the building, number of occupants and type of their activities, sprinklers, and other active firefighting measures should also be considered.

72

Fire Safety Design for Tall Buildings

3.11 FIRE PROTECTION METHOD A fire protection system offers a fire-resisting period—the time needed to evacuate the occupants of a building—that varies between 30, 60, 90 or 120 min. There are two main categories of fire protection methods: one is active control, and the other is passive control.

3.11.1 Active control system Active control system provides fire protections through the actions taken by a person or an automatic device such as sprinkler or firefighters, control smoke (Figure 3.17).

3.11.2 Passive control system The fire protection systems are built into the structural elements of a building such as intumescent paint, spray, and board protection on the structural steel members. 3.11.2.1 Intumescent paints An intumescent coating is basically made a paint-like material which is inert at low temperatures—but reacts with heat at high temperatures. As the temperature rises during a fire event, the intumescent coating swells

Figure 3.17 A typical sprinkler system in a room.

Fundamentals of fire and fire safety design

73

and forms a char layer that covers the steel. This char layer is of low thermal conductivity, thus acting as an insulating system. It should be noted that the coating usually expands up to 50 times when compared to its original thickness. The pros and cons of this type of protection are as follows: Advantages • Decorative finish • Rapid application • Easy-to-cover complex details • Easy post-protection • Fixings to steelwork, for example, service duct hangers • Quicker construction • Improved quality control • Reduction in site disruption • Cleaner sites • Improved site safety • Easier servicing installation. Limitations • Suitable for dry internal environments only • More expensive than spray • Require cleaned surfaces. 3.11.2.2 Spray fire protection In this method, a certain fire-resistive material is sprayed on the structural members on site. The appropriate thickness of the spray-applied fireresistive material is determined by a standard fire testing. Advantage • Fire-protective insulation can be applied by spraying to almost any type of steel member • Most products can achieve up to 4 h of protection • Low cost • Rapid application • Easy-to-cover complex details • Often applied to non-primed steelwork. Disadvantages • Appearance is poor for visible members • Overspray may need masking or shielding • Primer, if used, must be compatible.

74

Fire Safety Design for Tall Buildings

3.11.2.3 Board fire protection This method uses either gypsum board wrappings or insulation board enclosures on the structural members. The pros and cons of this type of protection are as follows: Advantages • Clean, dry fixing • Boxed appearance suitable for visible members • Guaranteed thickness. Disadvantages • Require careful fitting around complex detail • More expensive than sprayed protection.

3.11.3 Fire resistance test for protected members Similar to the fire test of unprotected member, the protected members can be placed into a fire furnace to heat up. The test procedure can follow ISO 834-11 (2014). In the meantime, it is loaded to failure. The time to failure can therefore be determined. Standard fire temperature will be used as well in the test. Two types of tests will be performed: Insulation tests These tests determine the thickness of fire protection needed to keep the average steel temperature at or below 550°C after a given fire resistance period. Stickability tests These tests ensure that the fire protection remains intact over the fire resistance period.

REFERENCES BS 476-20 (1987), Incorporating Amendment No. 1. Fire tests on building materials and structures, Part 20: Method for determination of the fire resistance of elements of construction (general principles). BSI ‘PD7974-3 (2019), Application of fire safety engineering principles to the design of buildings, Part 3: Structural response and fire spread beyond the enclosure of origin (Sub-system 3)’, The British Standards Institute. Ellobody, E., Bailey, C. G. (2011, June), Structural performance of a post-tensioned concrete floor during horizontally travelling fires. Engineering Structures, 33(6), pp. 1908–1917.

Fundamentals of fire and fire safety design

75

EN 1991-1-2 (2002), Eurocode 1. Actions on structures, Part 1-2: General actions. Actions on structures exposed to fire. Commission of the European communities. EN 1992-1-2 (2004), Eurocode 2. Design of concrete structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. EN 1993-1-2 (2005), Eurocode 3. Design of steel structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. EN 1994-1-2 (2005), Eurocode 4. Design of composite steel and concrete structures, Part 1-2: General rules. Structural fire design. Commission of the European communities. Fletcher, I., AHitchen, B., Welch, S. (2006), Performance of concrete in fire: A review of the state of the art, with a case study of the Windsor Tower fire, in: Proceedings of the 4th International Workshop in Structures in Fire, pp. 779–790. Fu, F. (2017). Grenfell Tower disaster: how did the fire spread so quickly? The Conversation. Horová, K., Tomáš, J., František, F. (2013, August–September), Temperature heterogeneity during travelling fire on experimental building. Advances in Engineering Software, 62–63, pp. 119–130. Ingberg, S. H. (1928), Tests of the severity of building fires. National Fire Protection Association, 22, pp. 43–61. ISO 834-11 (2014), Fire resistance tests—Elements of building construction, Part 11: Specific requirements for the assessment of fire protection to structural steel elements, ISO/TC 92/SC 2 Fire containment. ISO 834-2 (2019), Fire-resistance tests—Elements of building construction, Part 2: Requirements and recommendations for measuring furnace exposure on test samples, ISO/TC 92/SC 2 Fire containment. Lamont, S., Usmani, A. S., Gillie, M. (2004), Behaviour of a small composite steel frame structure in a “long-cool” and a “short-hot” fire. Fire Safety Journal, 39(5), p. 327. Law, M. A (1971), Relationship between fire grading and building design and contents. Fire Research Note Number 877. Fire Research Station, UK. NIST NCSTAR (2005, December), Federal building and fire safety investigation of the World Trade Center disaster, final report of the National Construction Safety Team on the collapses of the World Trade Center Towers. NIST NCSTAR 1A (2008), Final report on the collapse of world trade center building 7, National Institute of Standards and Technology, US Department of Commerce. Rein, G., Zhang, X., William, P., Hume, B., Heise, A., Jowsey Lane, A., Lane, B., Torero, J. L. (2007), Multi-storey fire analysis for high-rise buildings, in: Proceedings of the 11th International Interflam Conference, London, UK, pp. 605–616. Zhang, C., Li, G. Q., Usmani, A. (2013), Simulating the behavior of restrained steel beams to flame impingement from localized‐fires. Journal of Constructional Steel Research, 83, pp. 156–165.

Chapter 4

Structural fire design principles for tall buildings

4.1 INTRODUCTION In this chapter, the structural fire design principles will be introduced in depth on the basis of Chapter 3. It introduces the structural fire design procedures for steel, concrete, and composite structural members based on Eurocodes and British Standards. The two key structural fire design methods, namely, critical temperature method and moment capacity method, are both introduced. It also covers the design of post-tensioning slabs, connection, and beams with openings.

4.2 KEY TASKS FOR STRUCTURAL FIRE DESIGN As introduced in the preceding chapters, structural fire design is one of the key design tasks in fire safety design. This is because an effective structural fire design will assist in both evacuation route design and compartmentation design. The primary task for structural fire design is to design the fire resistance of structural (load-bearing) members such as beams, columns, wall, and slabs to ensure that they won’t fail within the required fire rating duration specified by various codes. In addition, the design shall ensure the structural stability of the whole building or local area of a building to avoid the entire or partial collapse of the building. Prevention of fire-induced collapse will be discussed in detail in Chapter 7.

77

78

Fire Safety Design for Tall Buildings

4.2.1 Building elements to be considered in design for fire In a tall building, building elements to be considered primarily in design for fire are as follows: • Structural frames (including structural beams and columns and the frame as a whole) • Floors (concrete floors or composite floors) • Load-bearing walls (such as shear walls and core walls) • Compartment walls (can be either load-bearing or non-load-bearing) Not included: Roofs—unless they are also designed to be part of means of escape.

4.2.2 Design of structural members in fire As introduced in Chapter 2, the structural members in fire can be designed in three domains: time domain, strength domain, or temperature domain. Two methods are recommended by the codes. One is in the temperature domain, called critical temperature method; the other is in the strength domain, called moment capacity method. They will be introduced in this chapter.

4.2.3 Design procedures In Chapter 3, the structural fire design steps specified in PD7974-3(2019) are introduced. The key steps can be further simplified as shown in Figure 4.1.

Figure 4.1 Key structural fire design procedure.

Structural fire design principles

79

Some of the design steps has been provided in Chapter 3; therefore, the last two steps will be introduced in detail in this chapter.

4.3 FIRE RESISTANCE RATING FOR LOAD-BEARING STRUCTURAL MEMBERS British Approved Document B (2019) Section 6, clause 6.1 specifies that “Elements such as structural frames, beams, columns, load-bearing walls (internal and external), floor structures and gallery structures should have, as a minimum, the fire resistance given in Appendix B, Table B3.” Fire resistance requirements for a structural member should be based on the parameters influencing fire growth and development. In addition, the type of structure and other design aspects such as evacuation and firefighting access also affect the fire resistance requirements of a structural member. The key factors that should be considered in the design include: 1. Factors affecting fire scenarios Probability of fire occurrence, fire spread, fire duration, fire load, severity of fire, ventilation conditions, fire compartment (type, size, geometry) 2. Type of the structural element Steel, concrete or composite, or timber 3. Factors affecting evacuation and firefighting • Evacuation conditions • Access for fire rescue teams Table B3 of Approved Document B (2019) is reproduced here in Table 4.1. It specifies minimum fire resistance time for major structural elements in a building.

4.4 DESIGN OF CONCRETE MEMBERS IN FIRE As already said, degradation starts when concrete is heated up. Significant deterioration will be noticed in all concrete members after the temperature rises up to 550°C, caused by the differential rate of thermal expansion of the constituent materials. The reinforcement responded differently from the plain concrete.

60 min or see Table B4 (whichever is greater) 30 min See Table B4

R see Table B4

REI 60 or see Table B4 (whichever is greater) R 30 and REI 15 REI see Table B4

2. Load-bearing wall (which is not also a wall described in any of the following items) 3. Floors

a. Between a shop and flat above

b. In upper storey of two-storey dwelling house (but not over garage or basement)

a. Any part a maximum of 1,000 mm from any point on the relevant boundary

REI see Table B4

See Table B4

See Table B4

REI see Table B4

b. Any roof that performs the function of a floor

5. External walls

30 min

REI 30

See Table B4

a. Any part forming an escape route

4. Roofs

Any other floor—including compartment floors

See Table B4

R see Table B4

1. Structural frame, beam, or column

Load-bearing capacity

Part of building minimum provisions when tested to the relevant European standard (minutes)

See Table B4

See Table B4

30 min

See Table B4

15 min

60 min or Table (whichever is greater)

Not applicable Not applicable

Integrity

See Table B4

See Table B4

30 min

See Table B4

15 min

60 min or see Table B4 (whichever is greater)

Not applicable Not applicable

Insulation

Alternative minimum provisions when tested to the relevant part of BS 476 (minutes)

Minimum provisions when tested to the relevant European standard (minutes)

Table 4.1 Fire resistance requirement for building members

(Continued )

Each side separately

From underside From underside

From underside

From underside

From underside

Exposed faces Each side separately

Type of exposure

80 Fire Safety Design for Tall Buildings

REI 60 or see Table B4 (whichever is less) REI see Table B4

b. Occupancies

c. Seven compartment walls (other than in item 6 or item 10).

See Table B4

REI 30 REI see Table B4

b. Any other part between the shaft and a protected lobby/corridor

c. Any part not described in (a) or (b) above

See Table B4

30 min

30 min

See Table B4

30 min

No provision

See Table B4

60 min or see Table B4 (whichever is less)

60 min or see Table B4 (whichever is less) See Table B4

60 min or see Table B4 (whichever is less)

No provision

60 min or see Table B4 (whichever is less)

30 min

Each side separately

Each side separately Each side separately

Each side separately

Each side separately

Each side separately

From inside the building

From inside the building

Type of exposure

Reproduced from Appendix B, Table B3 “Specific provisions of the test for fire resistance of elements of structure” of British HM Government Approved Document B (2019) Fire safety, In public domian on British HM Government website).

Not applicable 30 min

E 30

60 min or see Table B4 (whichever is less) 60 min or see Table B4 (whichever is less) See Table B4

a. Any glazing

8. Protected shafts excluding any firefighting shaft

REI 60 or see Table B4 (whichever is greater)

a. Flat from any other part of the building (see paragraph 7.1 of Approved Document B Volume 1)

6. Compartment walls separating either

30 min

RE 30

c. Any part beside an external escape route (Section 2 Diagram 2.7 of Approved Document B Volume 1 and Section 3, Diagram 3.4).

15 min

See Table B4

RE see Table B4 and REI 15

b. Any part a minimum of 1,000 mm from the relevant boundary

See Table B4

Alternative minimum provisions when tested to the relevant part of BS 476 (minutes) Load-bearing Integrity Insulation capacity

Minimum provisions when tested to the relevant European standard (minutes)

Part of building minimum provisions when tested to the relevant European standard (minutes)

Table 4.1 (Continued) Fire resistance requirement for building members

Structural fire design principles 81

82

Fire Safety Design for Tall Buildings

4.4.1 Thermal response of concrete in fire When exposed to fire, physical and chemical changes occur in concrete as follows: • At 100°C–140°C, evaporation of water starts. • At 300°C, the cement paste starts to shrink due to water evaporation. Aggregate starts to expand with spalling of concrete may be observed. • At 400°C–600°C, the calcium hydroxide in cement paste decomposes into calcium oxide and water, a significant reduction in strength. • At >550°C, the aggregate in the concrete start to decompose, a significant loss in concrete strength. The research from OFFSHORE TECHNOLOGY REPORT2001/074 shows that • Different curing method affects deterioration of reinforced concrete from around 60% to 85%. • Deterioration increased linearly slightly with the increasing rate of heating for the plain concrete but decreased nonlinearly when the concrete was reinforced. • There was a tendency for more severe deterioration for the plain concrete, but this was not statistically significant. • There was a faster increase in deterioration from plain lightweight aggregate concrete to normal concrete reinforced concrete.

4.4.2 Spalling During exposure to fire, heat and mass transfer happens in concrete actively. Under steep temperature gradient, the buildup of water pressure develops high local stresses, which may cause concrete spalling. The combination of thermal stress and the pore pressure due to the evaporation of the free water inside the concrete makes the spalling to happen. Zhukov (1975) developed a spalling model as shown in Figure 4.2; in Zhukov’s model, the stresses developed due to combined mechanical and thermal loads. He considered that the stresses acting could be categorized as load-induced stresses, thermal stresses, and pore pressures. 4.4.2.1 Types of spalling The two types of spalling are as follows: Explosive spalling occurs in a very early stage of a fire which is likely to lead to loss of protective cover of the main reinforcement. Rapid rises in temperature, such as short-hot fire, will result in strength loss

83

Load and thermal stresses

Moisture gain

Dry

Drying

Structural fire design principles

Temperature Compression Thermal stresses

Tension Pore pressure Vertical cracks Spall

Pore pressure Moisture

Load and thermal stresses

Figure 4.2 Zhukov’s spalling model (Khoury and Anderberg, 2000).

leading to reduction in fire resistance. The key factors affecting explosive spalling are the rate of temperature rise, the restraint conditions to thermal expansion, and the permeability of the concrete. Sloughing spalling, the concrete gradually comes away due to loss of effective bond and strength. 4.4.2.2 Prevention of spalling The explosive spalling usually occurs during the first 7–30 min of a fire incident. The spalling is unlikely to occur when the moisture content of the concrete is less than 3% by weight. EC2 EN 1992-1-2:2004/A1:2019 (E) specifies that explosive spalling shall be avoided but its influence on performance requirements (such as EI) shall be taken into account. Several preventive measures that need to be taken to achieve this requirement are introduced here. 4.4.2.2.1 Thermal barrier Thermal barriers usually limit the increase of temperature at the surface of the concrete. It includes fire-resistive materials—such as coating of fireproof paints and plastering of fire-proof mortars—that form a protective

84

Fire Safety Design for Tall Buildings

barrier on concrete structural member, control temperature rise in concrete surface layer, and thus reduce temperature gradient and the risk of spalling. 4.4.2.2.2 Reduction of vapor pressure The spalling happens primarily due to the evaporation pressure of the water; therefore, reducing the vapor pressure is an effective way of preventing the spalling. Several methods used for this are as follows: • Addition of polypropylene (PP) fiber • Forced drying of structural members • Installation of moisture eliminatory tubes. The most commonly used method in recent years is addition of PP fibers, which will be introduced here. 4.4.2.2.3 PP fiber PP fiber (as shown in Figure 4.3) has been widely used recently in some of the projects (primarily for tunnels) to resist concrete spalling in fire. A largescale fire test of European Concrete Building Project (ECBP) was conducted at BRE’s Cardington in 2001. The flat slab was supported by a number of high-strength (C85) concrete columns containing 2.7 kg/m³ PP fibers to reduce the tendency for explosive spalling. All of the columns survived the fire test without any significant spalling. The addition of fibers in concrete matrix bridges cracks and restrains them from further opening. Figure 4.4 is the scanning electron microscopic

Figure 4.3 Polypropylene fiber.

Structural fire design principles

85

Figure 4.4 SEM images of fracture of different fiber in coral concrete. The behavior of PP fiber. a) PP fibre inside concrete before fire. b) Channels left by PP fibre (Cai et al., 2020).

86

Fire Safety Design for Tall Buildings

image of the microstructure of PP fiber-mixed concrete under fire condition. It can be seen that PP fiber fills the void of the cement matrix. When temperature increases to 200°C, the PP fiber melts and leaves void channel (as shown in Figure 4.4). The channels can allow vapor to escape and thus reduce the buildup of high pore pressure within the concrete. Research by Cai et al. (2020a) shows that for an exposure to a high temperature of 600°C for 6 h, the loss in compressive strength is about half of that when no PP fibers are used. The lower contents of fibers generally showed worse performance due to further deteriorations and higher volumes of voids. As the content of PP increases, the slump of the mix decreases. Thus, for heavily reinforced concrete members, it is recommended to use super-plasticizers to enhance the workability.

4.4.3 Simplified calculation methods for concrete members from EC2 EN 1992-1-2:2004/ A1:2019 (E), 500°C isotherm method EC2 EN 1992-1-2:2004/A1(2019) specifies a simplified method to design structural members under bending and axial loads in fire. This method is valid for minimum width of cross section given in Table B1 of EC2 EN 1992-1-2:2004/A1:2019 and for a standard fire exposure depending on the fire resistance, as well as for a parametric fire exposure with an opening factor (O) of ≥0.14 m1/2 . This method is considering a reduction in the cross-sectional size of concrete beams due to heat damages on the concrete surfaces. The thickness of the damaged concrete is made equal to the average depth of the 500°C isotherm in the compression zone of the concrete section (Figures 4.5 and  4.6). This is a simplified method with simple assumption that concrete will lose all its bearing capacity when the member temperature rises to 500°C, while the residual concrete maintains its initial values of strength and elastic modulus. On the basic of the above assumptions, after reaching the 500°C isotherm, a new width bfi and a new effective height dfi of the section will be calculated by excluding the concrete outside the 500°C isotherm. The rounded corners of isotherms can be regarded by approximating the real form of the isotherm to a rectangle or a square. Then determine the reduced strength of the reinforcement according to the temperature of reinforcing bars in the tension and compression zones. Finally, the bending capacity of the beam post fire can be obtained based on the reduction factor of materials and the formula of bending capacity. Conventional calculation methods can then be used to determine the load-bearing capacity of the reduced cross-section.

Structural fire design principles

87

hfi

h

500ºC

T

bfi b

Figure 4.5 Reduced cross section of reinforced concrete beam and column with three-side fire exposure. (Reproduced from Figure B.1 of Page 72 BS EN 1992-1-2:2004+A1 (2019), Eurocode 2. Design of concrete structures, Parts 1–2: General rules. Structural fire design; permission to reproduce and derive extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: cservices@ bsigroup.com.)

500°C h fi

h

b fi b

Figure 4.6 Reduced cross section of reinforced concrete beam and column with four-side fire exposure. (Reproduced from Figure B.1 of Page 72 BS EN 1992-1-2:2004+A1 (2019), Eurocode 2. Design of concrete structures, Parts 1–2: General rules. Structural fire design; permission to reproduce and derive extracts from British and ISO standards is granted by BSI. British Standards can be obtained in PDF or hard copy formats from the BSI online shop: www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies only: Tel: +44 (0)20 8996 9001, Email: cservices@ bsigroup.com.)

88

Fire Safety Design for Tall Buildings

4.4.4 Concrete cover and protective layers Concrete cover can enhance the fire resistance of concrete members. Eurocode has a detailed requirement of minimum cover thickness for concrete members in fire conditions. In addition, as mentioned earlier, extra protective layers can also be used as a thermal barrier. The properties and performance of the materials used for protective layers should be assessed by appropriate test procedures.

4.5 DESIGN OF STEEL MEMBERS IN FIRE

4.5.1 Thermal response of steel in fire As shown in Chapter 3, the thermal and mechanical properties of steel change in elevated temperatures, and the load-bearing capacity of steel decreases dramatically. To be able to design a structural member in fire, Eurocode specifies a temperature domain design method, called critical temperature method. It will be introduced here.

4.5.2 The critical temperature method (BS5950, 2003 and EN 1993-1-2 2005) The critical (limiting) temperature of a member in a given situation depends on the load that the member carries under fire condition. Along with the increase of the fire, the limiting temperature is dependent on the fraction of the ultimate load capacity that a member has at the time of fire. The temperature causing the failure of a structural member depends on the utilization of the member in the fire situation. This is the simplest method of determining the fire resistance of a loaded member in fire conditions. However, when the load ratio is greater than 1, the member will fail at ambient temperature; this is because the structural member is overloaded. So, it failed primarily due to mechanical load rather than fire. 4.5.2.1 Assumptions Eurocode 4 specifies the following assumptions for using critical temperature method: • The temperature of the steel section is assumed to be uniform. • The method is applicable to symmetric sections with a maximum depth of 500 mm, and the slab thickness it supports is >120 mm.

Structural fire design principles

89

4.5.2.2 Load ratio (degree of utilization) In design codes such as British code BS 5950 Part 8 (BSI, 2003c) and European standards (EN 1991-1-2, 2004; EN 1993-1-2, 2005; EN 1994-1-2, 2005), the structural member at the fire limit state is assessed according to the level of load. It is called load ratio or degree of utilization. It is the resistance of applied load of the structural members at the time of fire compared to that at the ambient temperature. The concept of load ratio is the basis for the limiting temperature method for steel structures in BS 5950 Part 8 (BSI, 2003) and critical temperature method in Eurocode 3 (ENV 1993-1-2/4.2.4, 2005a). In BS5950 Part 8 (BSI, 1990), load ratio is defined as applied load (primarily due to dead load and live load) in fire conditions to those used in the design of the member at room temperature. In Eurocode 3 (ENV 1993-1-2 2/4.2.4, 2005), a similar ratio, which is called as degree of utilization is defined using the following formula: 1. For member Classes 1–3 and for all tension members:

μ0 =

Efi,d Rfi,d ,0

(4.1)

where Efi, d is the applied load under fire condition, R fi, d, 0 is the design moment of resistance of the member at ambient temperature. 2. Alternatively, for tension members and for beams where lateral-torsional buckling is not a potential failure mode, it is conservative to use: ⎡ γ M , fi ⎤ μ0 = ηfi ⎢ ⎥ ⎣ γ M0 ⎦

(4.2)

where 𝜂 fi, d is the reduction factor. 4.5.2.3 Critical temperature method for constrained members The critical temperature is the temperature at which failure is expected to occur in a structural steel element with a uniform temperature distribution. In Eurocode 3 (BS EN 3-1-2/4.2.4, 2005), the critical temperature is determined from ⎤ ⎡ 1 θ cr = 39.19ln ⎢ − 1⎥ + 482 3.833 0.9674 μ 0 ⎦ ⎣

(4.3)

90

Fire Safety Design for Tall Buildings

where μ0 is the degree of utilization as shown in Equation 4.1 (or called load ratio), It should be noticed that the above formula is primarily used for nonslender sections (Classes 1–3). For slender sections (Class 4), it is conservative to use 350°C as the critical temperature (as shown in Figure 4.7). According to Eurocode3 (EN 1993-1-2/4.2.4, 2005), this equation can be used only for member types for which deformation criteria or stability considerations do not have to be taken into account (such as beams). This allows its use for tension members and restrained beams but precludes its use for both columns and unrestrained beams, where stability phenomena must be considered. 4.5.2.4 Critical temperature method for the compression and unconstrained members Eurocode 3 (EN 1993-1-2/4.2.4, 2005) also provides the way to work out the critical temperature for compression members (such as columns) and unconstrained members, which is listed in Table 4.2. 4.5.2.4.1 Load ratio for column For columns in simple construction exposed up to four sides, the load ratio is as follows: U0 =

Ff Mfx Mfy + + Ag Pc Mb py + Zy Critical Temperature (C) 800 700

Class 1,2,3 sections

600 500 400 300

Class 4 sections

200 100 0

0.2

0.4 0.6 0.8 Degree of Utilisation

1

Figure 4.7 Relationship between critical temperature and degree of utilization for steel sections with different classes.

Structural fire design principles

91

Table 4.2 Critical temperature of steel compression members Compression member

Critical temperature (°C) for utilization factor (load ratio) 0.7

λθ i 0.4 0.6 0.8 1 1.2 1.4 1.6

485 470 451 434 422 415 411

0.6 526 518 510 505 502 500 500

0.5 562 554 546 541 538 536 535

0.4 598 590 583 577 573 572 571

0.3 646 637 627 619 614 611 610

0.2 694 686 678 672 668 666 665

Partially adapted from Eurocode BS EN1993-1-2/NA.2.6 (2005a).

where Ag is the gross area, pc is the compressive strength, py is the design strength of the steel, Zy is the elastic modulus about the minor axis, Mb is the lateral torsional buckling resistance moment, Ff is the axial load at the fire limit state, Mfx is the maximum moment about the major axis at the fire limit state, Mfy is the maximum moment about the minor axis at the fire limit state. For tension members exposed to four sides of fire, U0 =

Ff Mfx Mfy + + Ag Py Mcx Mcy

where Mcx is the moment capacity of the column about the major axis at the fire limit state, Mcy is the moment capacity about the minor axis at the fire limit state. 4.5.2.4.2 Column slenderness in fire In Eurocode 3, the column slenderness in fire can be expressed as follows:

λθ i = λ

ky , θ kE, θ

where λ is the column slenderness,

λ =λ

py L with λ = e 2 2 ry π E

92

Fire Safety Design for Tall Buildings

ky, θ is the reduction factor from Section 3 for the yield strength of steel at the steel temperature θ a reached at time t, k E, θ is the reduction factor from Section 3 for the slope of the linear elastic range at the steel temperature θ a reached at time t. 4.5.2.5 Column buckling resistance in fire The design of buckling resistance of column in fire a Class 1, Class 2, or Class 3 steel sections with a uniform temperature can be determined by N b,  fi ,  Rd =

2 fi Aky , θ f y   rM ,  fi

where 2fi is the reduction factor for flexural buckling in the fire design situation, ky, θ is the reduction factor from Section 3 for the yield strength of steel at the steel temperature ϴa reached at time t, 2fi is determined 2 fi =

1

ϕθ + ϕθ2 − λθ2

with ϕθ =

1 ⎡1 + αλθ + λθ2 ⎤ ⎦ 2⎣

and α = 0.65

235 fy

The axial compression force for column in fire can be calculated as follows: Pc ,  fi =

2 fi Pu,  fi 1.2

And Pu,  fi = Asky , θ py

4.5.3 Lateral torsional buckling of steel beams In certain circumstances, lateral torsional buckling needs to be considered for steel beams under fire condition. In Eurocode 3, lateral torsional buckling is calculated by the same equation at ambient temperature with inclusion of temperature effect.

Structural fire design principles

93

The beam slenderness in fire is given by

λLT , θ = λLT

ky , θ kE, θ

where λLT is the beam slenderness at ambient temperature. So, the lateral torsional buckling moment of beam under fire is Mb, θ =

2 LT , θ Mp, θ 1.2

where Mp, θ is the plastic bending resistance of the beam in fire. And 2 LT , θ =

1 2 ∅ LT , θ + ∅2LT , θ − λLT , θ

(

)

2 ⎤ ∅ LT , θ = 0.5 ⎡1 + α LT λLT , θ − 0.2 + λLT , θ   ⎣ ⎦

4.5.4 Beams in line with compartment walls When a beam is in line with a compartment wall, either above or supporting the wall, fire protection is required to limit deformation of the wall for the sake of integrity of the compartment (see Chapter 5). 4.6 MOMENT CAPACITY APPROACH (SECTION METHOD) The superseded BS 5950: Part 8 (2003) specifies a calculation method for finding the fire resistance of members in bending. It is called “moment capacity method” or “Section Method” in some literatures. Though the code was superseded, the proposed method has been widely used in some references (Cai, 2020b; Han et al., 2007; Xiang et al., 2010; Yang et al., 2002). This method is easy for design engineers to use. Therefore, it will be introduced in this section.

4.6.1 Method of calculation 4.6.1.1 Temperature profile The temperature distribution of the section at appropriate fire resistance periods must be known. Temperatures may be determined by fire tests or

94

Fire Safety Design for Tall Buildings

finite element modeling (as shown in Figure 4.8). Nowadays, it is primarily through heat transfer analysis available in most of the finite element analysis software. 4.6.1.2 Reduced strength of each element After temperature profile of the section is determined, the member is divided into a set of elements of approximately equal temperature (as shown in Figure 4.8). The number of elements will depend on the complexity of the member and its temperature gradient. The reduced strengths of the various elements of the cross section can be calculated at elevated temperatures. The strength reduction can be worked out using strength deduction factors from Eurocode, as it has been introduced in Chapter 3.

4.6.1.3 Reduced flexural strength calculation First, the “plastic” neutral axis of the section should be determined based on the reduced strength of each element. The moment capacity of the section then follows by multiplying the reduced strength of each element by the distance from the neutral axis and summing all the elements in the section.

467

460 467

518

456 465 465 456 478

478 605 663 690

460

467

467

Concrete

518

715

735

730 715

707

709

735

724 712 697 709

707

712 702 693 693 702 712 700

730 715

Figure 4.8 Element division and temperature profile of a typical composite beam section.

Structural fire design principles

P

2C14 3C25

1

8@150

200 150×8=1200

200 500

200×8=1600

150×8=1200

4000

8@150 3C25

8@150

8@200

200

2C14 400

1

P

400 200

548

200 500

95

250 1 1

Figure 4.9 The dimensions and reinforcement of a reinforced concrete beam.

4.6.2 Case study for flexural capacity of reinforced concrete beams using moment capacity approach In this section, based on the research of the author (Cai, Fu, et al., 2019), a case study of flexural capacity of reinforced concrete beams using moment capacity approach is shown here. As shown in Figure 4.9, a simply supported RC beam is exposed to fire. To obtain temperature profile of the RC beams, heat transfer analysis can be performed using ABAQUS. The specific heat and thermal conductivity, convection, radiation, and density of the materials should be first determined according to EN 1993-1-2 (2005), Eurocode 3. As shown in Figure 4.10, the temperature of concrete is obtained from the Abaqus model. The beam section is then divided as shown in Figure 4.11 with temperature Ti in the ith cell. The compressive strength reduction factor of concrete is obtained by substituting Ti into compression reduction factor of concrete in Equation 4.4. The temperature of the steel bar takes the highest temperature of the cell where the steel is located. The yield strength reduction factor from EN 1993-1-2 (2005), Eurocode 3 of reinforcement for the yield strength of the steel bar in high temperature is adopted.

Ψ cT =

∑Ψ

cTi

ΔbΔc

bxc

(4.4)

where b is the width of the beam section, xc is the height of the compression zone of the concrete beam section. The post-fire residual flexural capacity is determined using the moment capacity method. The height of the compressive zone (xc) of the concrete beam section after the fire can be calculated based on Equation 4.5: fc

∑Ψ

cTi

ΔbΔx + Ψ′yT f y′As′ = Ψ yT f y As

(4.5)

96

Fire Safety Design for Tall Buildings

Figure 4.10 ABAQUS simulation result of temperature profile of concrete section after 1 h of fire exposure.

m

Ƹx

Ƹb

Ă 1 2 3

h=m×Ƹx

xF

1 2 3 Ă b=n×Ƹb Figure 4.11 Section unit division of beam.

n

Structural fire design principles

97

where ΨcTi is the compressive strength reduction factor of the ith cell, Δb is the width of a single cell, Δx is the height of a single cell, Ψ′yT is the yield strength reduction factor of compressive steel, f′y is the yield strength of compressive steel at room temperature, N/mm², A′s and A s are the area of steel in the compressive and tensile zone, respectively, mm². After obtaining the height of the compression zone (xc), the residual flexural capacity (Mu) of the RC beams after fire can be obtained as follows: x ⎞ ⎛ Mu = α 1Ψ cT fc bxc ⎜ h0 − c ⎟ + Ψ′yT f y′As′ ( h0 − as′ ) ⎝ 2⎠ where α1 is the coefficient of equivalent rectangular stress figure in compression zone of concrete, which is 1 here; h0 is the effective height of the beam section; a′s is the distance from the resultant force point of the compressive steel reinforcement to the margins of the compressive section.

4.6.3 Flexural capacity of steel beams using moment capacity approach For designing steel member in fire, the code restricts the use of the moment capacity method to sections which are “compact” as defined in BS 5950: Part 1. It is assumed that the member fails in a flexural manner, without the occurrence of premature shear or instability effects in fire. In fire, the position of the “plastic” neutral axis changes when fire temperature changes. The lower parts of the cross section are normally exposed to the fire; therefore, the strength of lower parts is weakened, and hence the plastic neutral axis shifts upwards. For simple beams supporting concrete floors, the plastic neutral axis can rise close to the upper flange (as shown in Figure 4.12). In the moment capacity method, the strength reduction factor is applied to the original strength of the steel at different locations across the section on the basis of temperature distribution by finite element analysis. The section modulus of the beam in fire is therefore determined by considering the moment contribution of the stress blocks. It is normally found that the ratio of the section modulus of an I section in fire to that in normal design increases to a value between 1.2 and 1.5 depending on the proportions of the cross section.

98

Fire Safety Design for Tall Buildings

Concrete 0.83

Concrete 0.83

Plastic neutral axis 0.60 0.60 0.25 0.25

0.18 0.18

0.21

0.21

Figure 4.12 Position of plastic neutral axis and deducted stress distribution after fire.

4.7 DESIGN OF COMPOSITE BEAMS UNDER FIRE Composite beams are widely used in the steel-framed tall buildings. Figure 4.13 shows a typical steel composite building under construction. Composite floor systems have many forms such as composite beams with metal decking slabs (a most widely used in the construction). Figure 4.14 shows another type of composite beam which precasts hollow-core slabs as floor system.

4.7.1 Resistance of shear connection in fire Composite beams utilize the composite action between concrete slabs (in most cases metal decking slabs) and steel beams through shear connectors (as shown in Figure 4.15). Therefore, the behavior of shear connectors in fire will also affect the behavior of the composite action, and therefore the flexural capacity of composite beams in fire. The design formulas for shear resistance of shear studs in fire are given in EN 1994-1-2 (2005) Eurocode 4: Pfi, Rd = 0.8ku, θ PRd with PRd as obtained from Equation 6.18 of EN 1994-1-1. Pfi, Rd = kc, θ PRd with PRd as obtained from Equation 6.18 of EN 1994-1-1.

Structural fire design principles

99

Figure 4.13 A typical composite buildings. (Photo taken by the author.)

Figure 4.14 Composite beam with precast hollow-core slabs. (Photo taken by the author.)

100

Fire Safety Design for Tall Buildings

Figure 4.15 Shear connectors in composite beams.

where values of ku, θ and kc, θ are taken from Tables 3.2 and 3.3 of EN 1994-1-1, respectively. The temperature θ v 0C of the stud connectors and θ c 0C  of the concrete material may be taken as 80% and 40%, respectively, of the temperature of the upper flange of the beam.

4.7.2 Effect of degree of shear connection As it is widely known, the degree of shear connection greatly affects the capacity of the composite beams. BS 5950-8 (2003) specifies the effect of the different degree of shear interactions. Steel Construction Institute (SCI) P288 (2000) is represented in Table 4.3. Table 4.3 Limiting (Critical) temperature for composite beams with different degree of shear interactions Limiting temperature (°C) for a load ratio of

Degree of shear connection (%)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

100 40

485 525

520 550

550 575

580 600

610 635

645 665

685 700

SCI Publication P288 (2000), Fire Safety Design: A New Approach to Multi-Storey Steel Framed Buildings.

Structural fire design principles

101

4.7.3 Edge beams in fire As it is well known, the edge beam is designed as non-composite in most of the projects. But in fire design, the floor slab should be adequately anchored to the edge beam to ensure membrane action in the slab. Therefore, if it is designed to be non-composite, it must still provide shear connectors at minimum 300 mm spacing with steel mesh of the slab to be hooked to the shear connectors (SCI Publication P288, 2000). In addition, edge beams are supports for cladding, and to avoid fall of debris (such as broken claddings), the deformation during the fire should be controlled. Therefore, all the edge beams need to be fire protected. In addition, vertical ties and wind post can be used to further restrain the deformation.

4.7.4 Case study of composite beam design in fire The stress and strain distribution of a composite beam in fire conditions is illustrated in Figure 4.16. At the critical temperature, the lower flange of the beam is fully yielded. all of the steel section is in tension, making the plastic neutral axis of the composite section usually lie in the concrete slab. Therefore, in fire condition, higher bottom flange strains are generated. Therefore, a strain limit of 2% is recommended by SCI (1990) when assessing the moment capacity of the section, subject to the “stickability” of the fire protection if the member is fire protected.

4.8 DESIGN OF COMPOSITE SLABS IN FIRE Composite slab’s floor system has been widely used in tall buildings due to its good span to depth ratio. Figure 4.17 shows a typical metal decking composite floor spanning over a secondary beam. It is comprised of cast in situ concrete, steel mesh, shear connector, metal decking, and beams. Compression Concrete Deck Shear connector

Tension

In fire

Figure 4.16 Stress distribution of composite beams in fire (SCI, 1990).

102

Fire Safety Design for Tall Buildings

Cast-in-situ concrete

Steel mesh

Shear connector

Primary Beam

Metal decking Secondary Beam

Figure 4.17 A typical composite floor system.

All these components play important roles during the fire. The role of the shear connectors has been introduced in the last section. Metal decking also improves the integrity of the whole slab; in the meantime, it provides a shielding effect to the slab, shielding the heat flow into the concrete, and controls spalling.

4.8.1 Membrane actions in fire The results from Cardington tests show that composite slab plays a crucial role in fire. The behavior of composite slab in fire shows significant difference to its normal condition. In room temperature, they span in one direction along the metal decking and behave as a one-way slab. However, in fire, the slab acts as a membrane supported by the cooler perimeter beams and protected columns. Tensile membrane action is a load-bearing mechanism when slabs undergoing large vertical displacement, where the induced radial tension in the center of the slab is sustained by a peripheral ring of compression. A diagrammatic representation of tensile membrane action is provided in Figure 4.18. If the perimeter of the slab is fixed supported, compression membrane action starts first at small deflection stage. As the deformation becomes greater, tensile membrane action starts and loads are primarily carried by the reinforcement mesh. To enhance the tensile membrane action, the steel mesh in the slabs needs good anchorage to the supports.

4.8.2 Strength design composite slabs The moment capacity method can be used for the design of composite slabs under fire. The thermal profile shall be worked out, from which the reduced strength of the concrete and steel can be calculated. As a simplification of the actual behavior, the strength reduction factors may be taken according

Structural fire design principles

103

ne

n Zo

sio Ten

sion pres ) Com ("Ring" e Zon

Figure 4.18 Tensile membrane action in slab.

to the Eurocode. The moment capacities of the section in hogging, sagging, and bending can then be evaluated. Due to the complexity of working out the temperature and reduced strength of the metal decking, it is conservative to ignore the tensile capacity of the deck. Therefore, the calculation can be greatly simplified. For continuous floors, their capacities can be combined by considering the plastic failure mechanism of the floor. It is noticeable that there is a redistribution of moment during a fire from the mid-span area to the supports. 4.8.2.1 Calculation method based on plastic theory SCI Publication 056 (1991) introduces formulas for plastic failure of a continuous composite metal decking slabs in fire. For middle span Mh + Ms ≥ M0 where M h = hogging moment capacity in fire per unit width, M s = sagging moment of capacity in fire per unit width, M 0 = free bending moment per unit width. =

L2 (γ fd wd + γ fi wi ) 8

For end span ⎛ M ⎞ Mp + 0.5Mn ⎜ 1 − n ⎟ ≥ M0 ⎝ 8M0 ⎠

104

Fire Safety Design for Tall Buildings

where Mp = sagging, Mn = hogging moment capacity of the composite section, M 0 = free bending moment applied to the simply supported slab in fire conditions. The second term approximates to a value of 0.45 Mn. For an internal span of a continuous slab, the plastic moment capacity is simply Mn + Mp ≥ M0 4.8.2.2 Calculation method considering membrane action Bailey et al. (2000a, b) and Bailey (2001) developed a calculation method for determining the ultimate load-carrying capacity of two-way slabs incorporating the effects of tensile membrane enhancement under elevated temperatures. The design method calculates an enhancement factor due to effects of the membrane forces on the flexural strength. The method considers the failure mode shown in Figure 4.19. This method has been adopted by the SCI. In designing the slab, Bailey (2000) divides a composite floor into several horizontally unrestrained rectangular fire-resisting zones. These are composed internally of simply supported unprotected beams (Bailey and Moore, 2000a). With increasing exposure to elevated temperatures, the formation of plastic hinges in the unprotected beams redistributes the loads to the two-way bending slab, undergoing large vertical deflection. The design method also considers ultimate failure of the system based on the maximum permissible deflections due to the mechanical strains of the reinforcement and the thermal bowing deflections. Based on the diagram in Figure 4.20, the ultimate vertical deflection at the fire limit state is derived from a combination of thermal bowing of the slab and the mechanical strain in the reinforcement, which is defined in the following: v=

⎛ 0.5f y , θ ⎞ 3L2 α (T2 − T1 ) + ⎜ × 19.2h 8 ⎝ E, θ ⎟⎠ Fracture

Figure 4.19 Failure mode of slab in fire due to membrane action.

(4.16)

Structural fire design principles

Figure 4.20 In-plane stress distribution patterns by Bailey (2007).

105

106

Fire Safety Design for Tall Buildings

The deflection due to mechanical strain of the reinforcement is limited to span/30 (see Section 4.9.4 for further details). The composite slab capacity at any given time in fire is calculated as ⎛ Internal   work done by the composite slab in bending ⎞ w pθ = e ⎜ ⎝ External   work done by the applied  load per unit load ⎟⎠ ⎛ ⎞ Internal   work done by the beam in bending +⎜ ⎝ External   work done by the applied  load  per unit load ⎟⎠ where w pθ is the slab panel capacity at a given time, e is the enhancement of the slab capacity, calculated as in the reference (Bailey and Toh, 2007).

4.8.3 Insulation criterion of composite slabs BS 476: Part 20 (2014) specifies that the composite slabs should have the ability to limit the conduction of the heat to the upper surface. The average temperature raise of the upper surface should not exceed 140°C or a peak temperature of 180°C as shown in Figure 4.21. In determining temperatures Limit 140°C average rise

Top surface Temperature

100

100

150

150

200 250

200 250

300

300

400

400

500 600

500 600

700 800

700 800

Steel Mesh

Figure 4.21 A t ypical temperature profile of composite metal deck slab with the average temperature rise of the top surface 0) or G(R,S) > 0. This assurance is only possible in terms of the probability P(M > 0). This probability therefore represents a realistic measure of the reliability of the system or a structural member in fire. If the probability density functions f R(r) and fs(s) of R and S are available or can be approximated, and if R and S are statistically independent, the probability of failure Pf may can be expressed as follows: Pf = P(( R − S ≤ 0) =

∞ n≥ r

∫ ∫f

R

( r ) fs ( S ) drds

(6.26)

−∞ −∞

The reliability-based structural fire design and analysis are to work out Pf shown in Equation 6.26. The procedure for reliability-based design and analysis will be introduced in the following sections. 6.5.1.2 Reliability-based design and analysis procedure 6.5.1.2.1 Determination of limit state function The limit state function M = R − S needs to be determined first. This is based on the type of design problems (whether to design the flexural capacity of a beam under fire or the overall stability of a building in fire). 6.5.1.2.2 Monte Carlo simulation method The Monte Carlo technique is applicable for either stochastic or probabilistic problem. The process is computational and involves selecting input values at random for use in engineering calculations. Monte Carlo simulation is a choice of probability distributions for the random inputs. It uses the randomness to solve problems that might be deterministic in principle. Monte Carlo methods can be used to sample using a known probability distribution.

170

Fire Safety Design for Tall Buildings

It first needs to select a probability distribution for each individual variable. It is also essential to determine the dependencies between simulation inputs. Ideally, input data to a simulation should reflect what is known about dependence among the real quantities being modeled. Probability distributions for each variable need to be created from statistical data or information taken from real-life observations or experimental data. In the fire safety design, there are some key parameters that will affect the design values. For example, when determining atmosphere temperature, opening factor and fire load density are the two key parameters to affect its value, and they are mutually independent to each other. The probability distribution or the range of these parameters is readily known from design guidelines such as Eurocode and other research as shown in Table  6.1. Therefore, using the available distributions and key statistic index such as mean and standard deviation obtained from Eurocodes design practice (see Table 6.1), the random value of opening factor and fire load density can be generated. Subsequently, the corresponding atmosphere temperature can be calculated based on design formula. The random variables used are determined from large-scale data analysis and tests. 6.5.1.2.3 Determining statistical parameters of the variables The statistical parameters of design are given by Eurocode 1 (2002) and Eurocode 3 (2005). Using these statistic parameters or the range of the design parameters, the values of the variable can be generated using Monte Carlo simulation. Table 6.1 shows examples of statistical parameters of the variables used in the Monte Carlo simulation.

Table 6.1 Examples of statistical parameters of the variables used in the limit function Mean

Standard deviation

N/A

N/A

N/A

Gumbel

MJ/m2

420

126

Extreme type I

KN/m2

Variable

Distribution

Opening factor

Normal

Fire load density Imposed load

Units

Yield strength of steel

Log-normal MPa

280

28

Partial safety factors

Normal

-

-

-

Range

Source

0.02–0.2 Eurocode 1; Part 1.2 (2002) Eurocode 3 (2005) 1–5 Eurocode 1; Part 1.2 (2002) 275–355 Eurocode 1; Part 1.2 (2005) 1.5–2 Eurocode 1; Part 1.2 (2002)

Fire analysis and modeling

171

6.5.1.3 Case study for reliability analysis for individual members Cai and Fu et al. (2020) developed a new approach for post-fire reliability analysis of concrete beams retrofitted with CFRP sheet in bending using the Monte Carlo method. It is introduced here. 6.5.1.3.1 Limit state function The limit state function of a beam in bending in ambient temperature can be written as follows: Z = R − S = g ( X1 , X2 ,..., Xn )

(6.27)

where g(X) is the failure function, X1, X 2 , …, X n are n mutually independent random variables, R is the resistance of the structure, and S is the action effect of the structure. Values of Z greater than 0, less than 0, or equal to 0 indicate that the structure is under a reliable status, a failure status, or a limit status, respectively. The flexural capacity of RC beams at ambient temperature was (GB, 2010) as follows: MC = α 1fcbx ( h0 − 0.5x ) + f y′As′ ( h0 − as′ )

(6.28)

where MC is the flexural capacity of RC beams at normal temperature. With a random variable γm that represents the uncertainty coefficient of the resistance calculation, the limit state function of RC beams at normal temperature is as follows: Z = R − S = γ mMC − ( MGm + MQm )

(6.29)

where MGm is the mean value of MG and MQm is the mean of MQ. The limit state function of RC beams after fire is as follows: Z = R − S = γ m MCT − ( MGm + MQm )

(6.30)

Since the flexural capacity of RC beams was deteriorated after fire exposure, CFRPs can be used to reinforce the bottom of post-fire RC beams. If the bonding between CFRP and concrete is assumed to be perfect (GB, 2013), the flexural capacity of post-fire CFRP-reinforced RC beams can be computed as follows: MD = α 1ϕCT fcbx ( h − 0.5x ) + ϕ yT ′ f y′As′ ( h − as′ ) − ϕ yT f y As ( h − h0 ) x = (ϕ yT f y As + ψ f f fu,s Afe − ϕ yT ′ f y′As′ ) / (α 1ϕCT fc )

(6.31) (6.32)

172

Fire Safety Design for Tall Buildings

where M D is the flexural capacity of post-fire RC beams strengthened with CFRPs, f fu,s is the mean tensile strength of CFRPs, Afe is the valid sectional area of CFRPs, ψ f is the strength use coefficient of CFRPs. The limit state function of post-fire RC beams strengthened with CFRPs was determined as follows: Z = R − S = γ mMD − ( MGm + MQm )

(6.33)

6.5.1.3.2 Statistical parameters of the variables The statistical parameters are selected based on the Chinese Code (GB, 2010), EN 1992-1-2(2004)), and the research of Cai (2016) and Coile et al. (2014) (Table 6.2). 6.5.1.3.3 Monte Carlo simulation In the Monte Carlo simulation, the random variables for the limit state function were repeatedly simulated using the program coded in MATLAB, and the reliability can be calculated. The specific procedures are listed as follows: • The random variables of the limit state function were integrated with their probability distributions. • Random values were simulated repeatedly using the Monte Carlo method with the probability distribution of these random variables. • g(X) was calculated using the simulated values. • When the number of repetitions reached the preset value, the simulations were terminated. • Calculate the reliability based on the simulation results. 6.5.1.4 Case study for reliability analysis for a whole building To assess the probability of a whole building failure under fire is more complicated. Van Coile et al. (2014) developed a method considering the following probability: pig is the probability of a fire to develop, pf,u is the probability of early intervention by the occupants, pf,s is the probability active measures, such as sprinklers, pf,fb is the probability of the fire brigade, Pf,fi is the probability of the structure damage, Pf,1 is the probability of the structure fails.

Dead load Effective depth of section Beam width CFRP cross-sectional area CFRP tensile strength Total model uncertainty CFRP strip thickness T (°C) concrete compressive strength reduction factor T (°C) reinforcement yield stress reduction factor

SG h0 b Af ff γm tf

c

b

a

Bias: mean value/nominal value. CoV: coefficient of variation. std: standard deviation.

ϕ yT (ϕ yT ′ )

ϕCT

C30 compressive strength HRB335 steel yield stress Live load

Variable

fc fy SQ

Symbol

Beta

Lognormal Lognormal Extreme type I Normal Normal Normal Normal Weibull Normal Lognormal Beta

Distribution

-

-

mm

kN·m mm mm mm2 MPa -

MPa MPa kN·m

Units

Biasa (mean)

1.060 1.000 1.000 1.00 1.152 1 1.00 (Temperature-dependent, conforms to EN 1992-1-2) (Temperature-dependent, conforms to EN 1992-1-2)

1.395 1.139 0.859

Table 6.2 St atistical parameters of the variables used in the limit function

0.070 0.030 0.010 0.02 0.08 0.025 0.010 (Temperature dependent) (Temperature dependent)

0.15 0.07 0.233

CoVb (stdc)

-

565 250 3,100 0.167 -

20.1 335 -

Nominal value Source

GB (2012) CAI et al. (2016) CAI et al. (2016) CAI et al. (2016) CAI et al. (2016) Coile et al. (2014) CAI et al. (2016) EN 1992-1-2:2004 EN 1992-1-2:2004

GB (2010) GB (2010) GB (2012)

Fire analysis and modeling 173

174

Fire Safety Design for Tall Buildings

This analysis process proposed by Van Coile et al. (2014) is shown in Figure 6.14. It comprises two domains: the ‘event instigation’ and ‘response’ domains. Based on this framework, Hopkin et al. (2017) made a reliability analysis of the probability of failure of a tall building in fire. The Monte Carlo simulation is used to sample different variables, such as fire load density and temperature spread rate; therefore, different fire scenarios can be simulated and the probability of failure of the building can be assessed.

6.5.2 Fire fragility functions Fragility function has been recently used by researchers to characterize the probabilistic vulnerability of buildings to fire. In earthquake engineering, fragility functions are widely used to assess the likelihood of structural damage due to an earthquake. A fragility function provides the probability of exceeding a damage state for a given intensity of earthquake load. Similarly, fire fragility functions can be developed to measure the probability of exceeding a damage state (e.g. column failure, excessive beam deflection) for a given intensity of fire load. This method is quite new, and the influence of the different uncertain parameters on the functions has not been systematically studied. Uncertainties in fire, heat transfer and structural models, fire load intensity, compartment geometry and openings, the thickness and thermal conductivity of fire protection, and the material degradation, all generate significant variability in the fire fragility. The prevailing parameters in constructing fire fragility functions for steel frame buildings identified through sensitivity analyses are conducted using the Monte Carlo simulations and a Fire ignition in building

Occupants fail to control the fire

,

Sprinklers fail to operate successfully

,

Structural fails when subject to a fully developed fire

AND

Fire becomes fullydeveloped =

Fire & rescue service intervention is unsuccessful Event instigation domain

,

+

,

AND +

,

+

,

Probability of fire induced structural failure in one year reference period ,

=

,

+

Fully developed fire and structural response domain

Figure 6.14 Stochastic factors leading to a fire-induced structural failure (Coile et al., 2014).

Fire analysis and modeling

175

variance-based method. One important parameter in defining a fragility function is intensity measure for fire load. Gernay et al. (2019) used average fire load (in MJ/m² of floor area) as the intensity measure. 6.5.2.3 Compartment-level fragility function Khorasan et al. (2016) and, Gernay et al. (2019) developed the below equation: ∞

PF⎜Hfi =

∫ ⎡⎣1 − F

D⎜Hfi

(α )⎤⎦ fc (α ) dα

(6.34)

0

where PF│Hfi is the probability of reaching a damage state condition to the occurrence of a fire Hfi, FD│Hfi (.) is the cumulative distribution function of the demand relative to the fire Hfi, fC (·) is the probabilistic distribution function of capacity. PF│Hfi is obtained through repeated structural fire analysis under fire load densities (q values) in the same compartment. The analysis needs to be repeated sufficient times to be able to work out PF│Hfi. Based on PF│Hfi, the fragility function is built by fitting a function to the obtained points, assuming a lognormal distribution: ⎡ ⎛ q⎞ ⎤ ⎢ ln ⎝⎜ c ⎠⎟ ⎥ ⎥ F (q ) = ∅ ⎢ ⎢ ζ ⎥ ⎢⎣ ⎥⎦

(6.35)

where q is the fire load (MJ/m²), Φ[•] is the standardized lognormal distribution function. There are two parameters c and ζ: c is the mean of lognormal distribution, ζ is the standard deviation of the lognormal distribution. c and ζ are determined by the best fit to the data points from structural fire analysis. 6.5.2.4 Building-level fragility function Fire fragility functions should first be developed for each compartment under different fire scenarios and then combined to derive a fire fragility function for the entire building. The combined fragility function is also a

176

Fire Safety Design for Tall Buildings

lognormal function, the same as in Equation 6.35. Khorasan et al. (2016) and Gernay et al. (2019) developed the following equations to work out the two lognormal parameters: n

qc =

∏c

pi i

(6.36)

i=1

ζ c2 = PT Z + AT QA ⎡ p (1 − p1 ) ⎢ 1 Q= ⎢  ⎢ − p n p1 ⎢ ⎣

(6.37) 

− p1pn

 

 pn (1 − pn )

⎤ ⎥ ⎥ ⎥ ⎥ ⎦

(6.38)

where qc is the mean of combined lognormal distribution, ζc is the standard deviation of the combined lognormal distribution, Other parameters are as follows: n is the number of ‘nominally identical but statistically different’ fragility curves, ci is the median associated with each individual fragility curve, pi is the conditional probability for a fire in compartment, i, a fire occurs in the building, P is the vector of the probabilities pi, Z is the vector of the variances, ζi ² is associated with each individual fragility function, A is the vector of the expected values (ln ci), and ⎡ p (1 − p1 ) ⎢ 1 Q is the matrix given by Q = ⎢  ⎢ − pn p1 ⎢ ⎣



− p1pn

 

 pn (1 − pn )

⎤ ⎥ ⎥ ⎥ ⎥ ⎦

Based on the above approaches, fire fragility curve can be derived. Figure 6.15 shows typical fire fragility curves for different types of buildings. 6.5.3 Other probabilistic approaches in fire safety design Fu (2020) used the Monte Carlo simulation to simulate a probability distribution for different variables for the random inputs for fire-induced tall building collapse using machine learning. This will be introduced in detail in Chapter 9.

Fire analysis and modeling

177

Figure 6.15 Fire fragility curves for different types of buildings.

6.6 MAJOR FIRE ANALYSIS SOFTWARE There are several fire analysis software packages available on the market for engineers to choose. Research-focused software developed by certain academic research groups, including VULCAN, ADAPTIC, and SAFIR, focus on highly specific modeling issues. They are not widely used by engineers, due to their availability. Conversely, many commercial software packages, including DIANA, ANSYS, and ABAQUS, are widely used by engineers, but the source code is proprietary and expensive to purchase. Some open-source codes such as OpenSees can be freely accessed, and users can develop the software package based on their needs. It also allowed for the modification of stiffness matrices and load vectors, as well as develop new element. This is usually difficult in commercial software. In terms of fire, it has a library of temperature-dependent materials and thermal elements for both shells and beams. It is also able to perform thermalmechanical analysis. ABAQUS and ANSYS have been introduced in the preceding sections of this chapter; therefore, in this section, other major fire analysis software will be introduced.

6.6.1 Ozone software Ozone software is developed by the University of Liege, Belgium. It was developed primary based on the theory of zone models. It includes a twozone model and a one-zone model with a possible switch from two to one

178

Fire Safety Design for Tall Buildings

zone if some criteria are encountered. Therefore, it can model both localized and fully developed fires. It can also solve the member temperatures, such as wall model which is made by the implicit finite element method. Different combustion models have been developed to cover different situations of use of the code. A Graphic User Interface has been developed to define the input data.

6.6.2 CFAST CFAST is a free and open-source software provided by the National Institute of Standards and Technology (NIST) of the United States Department of Commerce. CFAST is another two-zone fire model capable of predicting the environment in a multi-compartment structure subjected to a fire. It calculates the time-evolving distribution of smoke and gaseous combustion products as well as the temperature throughout a building during a userprescribed fire.

6.6.3 FDS Fire Dynamic Simulator (FDS) is a large-eddy simulation (LES) code for low-speed flows, with an emphasis on smoke and heat transport from fires. It’s one of the most used software to study fire dynamics. The Fire Structure Interface (FSI) module in FDS can be used to impose the gas temperatures from the FDS simulations. It is used to predict the evolving thermal state of the building WTC7 (NIST NCSTAR, 2008).

6.6.4 LS-DYNA LS-DYNA was capable of explicitly modeling failures, falling debris, and debris impact on other structural components. It could also model nonlinear and dynamic processes, including nonlinear material properties, nonlinear geometric deformations, material failures, contact between the collapsing structural components, and element erosion based on a defined failure criterion. In addition, LS-DYNA had capabilities to include thermal expansion and softening of materials. Therefore, it is used by NIST in their investigation on the World Trade Centre 7 collapse (Figure 6.16).

6.6.5 OpenSees OpenSees (Open System for Earthquake Engineering Systems) is an open-source software, initially developed by the University of California in collaboration with PEER (the Pacific Earthquake Engineering Research Centre) and NEES (Network for Earthquake Engineering Simulation). OpenSees has a sizeable community of developers and users who are

Fire analysis and modeling

179

Figure 6.16 Use of LS_DYNA to simulate fire-induced collapse of WTC7. (Reprint from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center disaster. Final report on the collapse of the World Trade Center Towers. (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1A, Figures 3–11, p. 40. https://doi.org/10.6028/NIST.NCSTAR.1a.)

devoted to advancing the toolkit available to structural engineers. Although the software was first developed as a tool for assisting in earthquake analysis, it has been extended in recent years and, as such, has emerged as a valuable instrument that structural engineers can use to analyze nonlinear structural responses. A long-term project was initiated at the University of Edinburgh in 2009 to extend OpenSees’ capabilities to the analysis of fire, heat transfer, and thermo-mechanical analyses (Jiang et al., 2012). This involved the establishment of material models at increased temperatures based on the Eurocodes along with thermal load classes. Additionally, to examine thermal effects, temperature-dependent designs have been included for simple element types, including shell and beam elements (Jiang et al., 2012). The material library associated with the early version of OpenSees has also been extended by including temperature-dependent material models for concrete and steel, which are again consistent with the Eurocodes (Jiang et al., 2012). Thermo-mechanical beam-column elements are derived from the base class element, and keep the general interface and data structure from the beamcolumn elements for ambient temperature use, such as forming the stiffness matrix and residual forces (Jiang, 2012).

180

Fire Safety Design for Tall Buildings

REFERENCES BRE (2004), ‘Client report: Results and observations from full-scale fire test at BRE Cardington, 16 January 2003 Client report number 215–741’, February 2004. (Accessible from: http://www.mace.manchester.ac.uk/project/research /structures/strucfire/DataBase/TestData/default1.htm) BS 476-20 (1987), Incorporating Amendment No. 1. Fire tests on building materials and structures —Part 20: Method for determination of the fire resistance of elements of construction (general principles) Cai, B., Zhang, B. and Fu, F. (2020). Post-fire reliability analysis of concrete beams retrofitted with CFRPs: a new approach. Proceedings of the Institution of Civil Engineers - Structures and Buildings, 173(11), pp. 888–902 Coile, R. V., Caspeele, R., Taerwe, L. (2014), Reliability-based evaluation of the inherent safety presumptions in common fire safety design. Engineering Structures, 77, pp. 181–192. Elhami Khorasani, N., Gernay, T., Garloc, M. (2016), Fire fragility functions for community resilience assessment. Proceedings of the 9th International Conference on Structures in Fire, 8–10 June 2016, Princeton University, NJ. EN 1990:2002+A1 (2005), Eurocode. Basis of structural design, Commission of the European communities. EN 1991-1-2 (2002), Eurocode 1. Actions on Structures, Part 1–2: General actions. Actions on structures exposed to fire. Commission of the European communities. EN 1993-1-2 (2004), Eurocode 3. Design of steel structures, Part 1–2: General rules – Structural fire design. BSI, London, UK EN 1993-1-2 (2005), Eurocode 3. Design of steel structures, Part 1–2: General rules. Structural fire design. Commission of the European communities. EN 1994-1-2 (2005), Eurocode 4. Design of composite steel and concrete structures, Part 1–2: General rules. Structural fire design. Commission of the European communities. Fu, F. (2015), Advanced Modeling Techniques in Structural Design. John Wiley & Sons, Ltd. ISBN 978-1-118-82543-3. Fu, F. (2016a), Structural Analysis and Design to Prevent Disproportionate Collapse. CRC Press. ISBN 978-1-4987-8820-5. Fu, F. (2016b), 3D finite element analysis of the whole-building behavior of tall building in fire. Advances in Computational Design, 1(4), pp. 329–344 Fu, F. (2020), Fire induced progressive collapse potential assessment of steel framed buildings using machine learning. Journal of Constructional Steel Research, 166, pp. 105918–105918. GB (2010), GB 50010-2010: Code for design of concrete structures. GB, China, Beijing. GB (2012), GB 50009-2012: Load code for the design of building structures. GB, China, Beijing. GB (2013), GB 50367-2013: Code for design of strengthening concrete structure. GB, China, Beijing. Gernay, T., Khorasani, N. E., Garlock, M. (2019), Fire fragility functions for steel frame buildings: Sensitivity analysis and reliability framework. Fire Technology, 55(4), pp. 1175–1210. DOI: 10.1007/S10694–018–0764–5

Fire analysis and modeling

181

Hopkin, D., Anastasov, S., Swinburne, K., Lay, S., McColl, B., Rush, D., Van Coile, R. (2017), Applicability of ambient temperature reliability targets for appraising structures exposed to fire. CONFAB 2017 Conference Proceedings. Jiang J (2012), Nonlinear thermomechanical analysis of structures using OpenSees. PhD Thesis, University of Edingburgh, Scotland. NIST (National Institute of Standards and Technology) NCSTAR (National Construction Safety Team) (2008), Federal building and fire safety investigation of the World Trade Center disaster. Final report on the collapse of World Trade Center Building 7. NIST, U.S. Department of Commerce, Gaithersburg, MD. Novozhilov, V. (2001), Computational fluid dynamics modelling of compartment fire. Progress in Energy and Combustion Science, 27(6), pp. 611–666. Purkiss, J. A. (2007), Fire Safety Engineering Design of Structures, ButterworthHeinemann, Elsevier, Linacre House, Jordan Hill, Oxford, UK Van Coile, R., Balomenos, G. P., Pandey, M. D., Caspeele, R. (2017), An unbiased method for probabilistic fire safety engineering, requiring a limited number of model evaluations. Fire Technology, 53, pp. 1705–1744. Cai B., Zhao L. L. and Yuan Y. H. (2016) Reliability of bending capacity for corroded reinforced concrete beam strengthened with CFRP. Concrete 10, 148–151.

Chapter 7

Preventing fire-induced collapse of tall buildings

7.1 INTRODUCTION In this chapter, how to design a building to prevent fire-induced collapse will be discussed. The collapse mechanism of a building in fire and methods for mitigating the collapse of a tall building will be introduced, all based on the existing research and fire-induced collapse incidents. As steel-framed buildings are more vulnerable in terms of fire-induced collapse, this type of buildings will be focused in this chapter.

7.2 DESIGN OBJECTIVE AND FUNCTIONAL REQUIREMENT FOR STRUCTURAL STABILITY IN FIRE Although the first priority of a fire safety design is to save lives of occupants rather than preventing the collapse of buildings (Fu, 2016a), any collapse of the buildings will cause huge economic loss. In addition, collapse of a building can also have a heavy impact on life safety design target. Therefore, building regulation of New Zealand, building Act 1991, specifies a design to have the following requirements:

C4.1 STRUCTURAL STABILITY DURING FIRE OBJECTIVE • Safeguard people from injury due to loss of structural stability during fire • Protect household units and other property from damage due to structural instability caused by the fire.

183

184

Fire Safety Design for Tall Buildings

C4.2 FUNCTIONAL REQUIREMENT • Buildings shall be constructed to maintain its structural stability during fire to • Allow people adequate time to evacuate safely • Allow fire service personnel adequate time to undertake rescue and firefighting operations • Avoid collapse and consequential damage to adjacent house units or other properties. 7.3 IMPORTANCE OF COLLAPSE PREVENTION OF TALL BUILDINGS IN FIRE From Section 7.2, it can be seen that in a fire safety design, the primary purpose of a building’s structural stability is to save the lives of occupants in the event of fire, rather than preventing collapse Currently, the major objective of structural fire design in most design codes is to ensure loadbearing capacity of the building to continue to function until successful evacuation of the occupants, rather than prevent collapse of the building. Therefore, thus far, there is no clear guidance for designing a building to prevent fire-induced collapse across the world. However, as introduced in Chapter 1, there are several incidents of fireinduced collapse of tall buildings such as World Trade Center (WTC1, WTC2, and WTC7). As explained in Section 7.2, even for the purpose of saving lives of the occupants, preventing or delaying the collapse of a building in fire is also essential in a fire safety design.

7.4 COLLAPSE MECHANISM OF TALL BUILDINGS IN FIRE All the collapse of a building starts from a local failure of structural members. As introduced in Chapter 2, there are four major failure modes of structural members in fire discovered by Cardington tests: • • • •

Beam buckling and yielding Column buckling and yielding Connection failure Slab failure.

These structural members are not working independently; they have impacts on each other. Under thermal expansion and subsequent contraction in the cooling stage, the interaction between the structural members causes extra stress and deformation to them, which makes it difficult to

Preventing fire-induced collapse

185

predict the actual failure mode of the whole building. However, as introduced in Chapter 2, it is found from WTC1 and WTC7 collapse incidents that the column buckling is the key reason for the collapse of these two buildings. The column buckling will trigger the failure of the floor above. The floor failure in both WTC1 and WTC7 caused further failure of surrounding columns in the horizontal direction and triggered progressive failure of the whole building. Therefore, it is worth investigating the response of individual structural members first.

7.4.1 Factors affecting thermal response and failure mechanism of individual members The thermal behavior of structural members in fire is a complex problem, and the response of individual members in a tall building subjected to fire loading is affected by the following parameters: 1. 2. 3. 4.

Temperature profile of the member Degree of thermal restraint offered by the surrounding members Degradation in material properties with increasing temperature Capacity of deployment of alternative load-carrying paths for adjacent members.

To better understand the behavior of the structural member in fire, a 3D finite element model of a typical single-storey composite frame is set up in Abaqus® by the author, as shown in Figure 7.1. To facilitate further discussion on the modeling result, this model is designated as Model 1 in this chapter.

Figure 7.1 Model 1 of a typical composite floor in Abaqus®. (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

186

Fire Safety Design for Tall Buildings

In the analysis, parametric fire temperature is used for gas temperature, Short-hot fire scenario is adopted. Figure 7.1 shows the temperature distribution of slabs, beams, and columns after analysis. The edge columns and beams and columns are fire protected. Therefore, it can be seen that the temperature of the slabs is higher than that of beams and columns due to their fire protection.

7.4.2 Behavior and failure mechanism of steel beams in fire 7.4.2.1 Local buckling of beams in connection area In the event of fire, the steel beams experience expansion at the heating stage, but this expansion is restrained by the connections and columns which they are connected to. To clearly understand the behavior of the beams in a fire condition, Model 1 is used here to demonstrate the change of internal force in the beams. As shown in Figure 7.2, the axial force of the beam first increases due to thermal expansion; however, due to the restrain of the connections and columns, the tensile force starts to drop at around 300 s, as the constrain causes the compression force in the beams. In Cardington tests, all primary beams showed signs of local buckling near to the connections. All the secondary beams also exhibited local buckling at the connection zone as well. However, local buckling did not occur where beams were supported on edge beams. This is due to the buckling stresses being relieved by rotation of the edge beam at the connection. In most of the beams, lower flange buckling was observed. The reason for local buckling at connection zone is due to the constraint of the connection to the beam during the expansion in the heating stage. 4500000 4000000

Axial Force (N)

3500000 3000000 2500000 2000000 1500000 1000000 500000 0 -100

0

100

200

Time (S)

Figure 7.2 Tensile axial forces of beams.

300

400

500

Preventing fire-induced collapse

187

This constraint causes plastic compression deformation of the beam especially in the lower flange. This is clearly demonstrated in in Figure 7.3. It can be seen that, the plastic strain developed in the beam at connection location during the fire in Model 1. It can be seen that at around 200 s, plastic strain starts to develop. 7.4.2.2 Excessive deflection In all Cardington tests, the unprotected steel beams were severely deformed, at a beam temperature of 500°C–600°C. As shown in Figure 7.4, the deflection of the beam increases when the temperature increases. 0.018 0.016 0.014

Plastic Strain

0.012 0.01 0.008 0.006 0.004 0.002 0 -100

-0.002

0

100

200

300

400

500

Time(s)

Figure 7.3 Plastic stain development of the beam at connection location.

Verticalal Deflection (m)

0.03 0.025 0.02 0.015 0.01 0.005 0 -100

0 -0.005

100

200

Time(s)

Figure 7.4 Vertical deflection of beam at middle span.

300

400

500

188

Fire Safety Design for Tall Buildings

7.4.3 Behavior and failure mechanism of slabs in fire Figure 7.5 shows the deflected shape of a typical steel composite floor in fire. As introduced in Chapter 3, Section 3.4.3, the slab was heated at a much slower rate than the steel beam due to its greater thermal inertia. It has a greater thermal gradient compared to steel beams, with a lower mean slab temperature. Initially, the slab response was driven largely by the behavior of the rapidly expanding steel beams. This was due to their large compressive forces to the corresponding post-buckling deflections of the composite beam and to the consequent large moments due to a P-δ effect. However, after the steel beams could no longer transfer a force to the slab, the stress state within the slab itself drove the response of the floor system as a whole (O’Connor et al., 2003). 7.4.3.1 Membrane actions of slabs At ambient temperature, the failure mode of slabs follows patterns of yield lines. However, based on the observations of the BRE Cardington tests and other fire tests on tensile membrane action, mode of failure and the design method are suggested by Bailey and Moore (2000a,b) as introduced in Chapter 5. In Cardington tests, despite failures of steel frame, the floor slab continued to resist loading due to membrane action developed in the slab. Damage to the floor slab steel mesh was found. In all tests, concrete cracks were observed along compartment boundaries or the main primary and secondary beams. The most extensive damage to the concrete slab was observed around the central column. This damage occurred at the cooling stage due

Figure 7.5 Abaqus® model showing deflected shape of a typical steel composite floor. (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

Preventing fire-induced collapse

189

Figure 7.6 Model 2, a tall building under long-cool fire in storeys 9–11. (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

to the high forces generated by differential cooling period between concrete and steel structures. To tackle this failure mode, sufficient anchorage of reinforcing mesh at the boundary is essential in keeping the integrity of the compartment as well as resisting potential collapse of the buildings. Figure 7.6 shows an Abaqus® finite element model set up by the author to simulate a tall building under long-cool fire. It is denoted as Model 2 in this chapter. The model replicates a 20-storey steel-framed tall building with cross-bracing in four sides of the building as major lateral stability system. It has 7 × 7 bays at each direction. The fire was set in storeys 9–11. Figure 7.7 is plastic strain distribution in the floor plate at storey 11. It can be seen that plastic strain developed at each bay following the pattern of membrane action. The floor slabs resisted the loads by tensile membrane action and formed a tensile zone in the middle of the fire-exposed slab, surrounded by a compression ring. It is recommended that fire protection will provide the necessary perimeter vertical support along the slab panel boundaries to ensure membrane action and, therefore, resist the collapse of the building. 7.4.3.2 Effect of different fire scenarios in composite slabs As discussed in Section 3.4.4 of Chapter 3, the research by Lamont et al. (2004) discovered that the stress state of the composite beams is significantly affected by the heating regime. The most detrimental fire scenario

190

Fire Safety Design for Tall Buildings

Figure 7.7 Principle plastic strain distribution of floor plate at storey 11. (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

for composite beams is the “short-hot” fire. Because the temperature gradient is high, greater thermal bowing effect, causing a greater tension in the composite section and large tensile strains, is observed in the slabs. Long-cool fire results in higher temperatures in the concrete and the protected steel. This results in greater displacements in the protected structural elements but much later in terms of real time. However, because the concrete slab achieves higher temperatures, there is much less tension in the slab with growing compression towards the end of heating. 7.4.3.3 Other research in composite slabs in fire Wong and Burgess (2013) found that the vertical supports provided by protected beams along the edge of the slab are important in the development of the tensile membrane actions. Lin et al. (2015) investigated the effect of protected beams on the fire resistance of composite buildings. The results showed that non-fixed vertical support significantly reduced the development of tensile membrane action. In comparison to the case with fixed support, tensile membrane action was fully mobilized. Jiang and Li (2014) studied parameters affecting tensile membrane action of reinforced concrete floors in fire. It was found that failure modes of the slab depend on reinforcement layout, aspect ratio, and boundary condition. Nguyen and Tan (2017) conducted experiments on three one-quarter scale composite slabs with different bending stiffnesses of protected edge beams under fire conditions. The results showed that an increase in the edge beam-bending stiffness initially reduced the deflection. At higher temperature, the effect of greater stiffness of the edge beams was negligible.

Preventing fire-induced collapse

191

All of the above studies confirmed that protected edge beams have a significant effect on the fire resistance of the structure.

7.4.4 Behavior and failure mechanism of steel column in fire 7.4.4.1 Change of column force in fire Figures 7.8–7.10 show the development of axial compression force of columns at different locations in Model 2. It can be seen that due to thermal expansion, the axial compression force in corner and edge column shows

Axial force in coulmn (N)

0 -500 0 -200000

500

1000

1500

2000

2500

3000

-400000 -600000 -800000 -1000000 -1200000 -1400000 -1600000 -1800000

Time (S)

Figure 7.8 Compression axial forces of a perimeter column at level 10 during fire.

0

Axial force in coulmn (N)

-500

0

500

1000

1500

2000

2500

3000

-500000 -1000000 -1500000 -2000000 -2500000 -3000000

Time (S)

Figure 7.9 Compression axial forces of an internal column at level 10 during fire.

192

Fire Safety Design for Tall Buildings

Axial force in coulmn (N)

0 -500 0 -100000

500

1000

1500

2000

2500

3000

-200000 -300000 -400000 -500000 -600000 -700000 -800000 -900000

Time (S)

Figure 7.10 Compression axial forces of a corner column at level 10 during fire.

certain percentage of decrease when the temperature of fire rises. The corner columns show a greater drop of compression force than the edge columns. However, the axial forces in internal columns show an increase. The reason for these force distributions can be explained by Figure 7.11, which shows the overall deformed shape of the entire building in fire. It can be seen that due to the thermal expansion, the edge and corner columns are

Figure 7.11 Deform shape of Model 2 in fire (deformation amplified 15 times). (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

Preventing fire-induced collapse

193

deforming outward, therefore making their force reduce and increasing the column force in internal columns. 7.4.4.2 Out plane bending of columns To better understand the behavior of the columns, the Cardington test is simulated using ANSYS as shown in Figure 7.12. Fire was set at storey 4 in the corner to simulate Conner fire test in the Cardington test (as shown in Figure 7.12, which shows temperature contour of the building elements). This is different to the fire scenarios in Model 2, where all three floors are set to fire. Figure 7.13 shows the moment of the columns on the third floor under fire loading. A huge bending moment was observed. Due to the effects of material degradation, the bearing capacity of these columns under fire is significantly reduced, so the buckling load reduces. As we know, buckling of the columns is one of the major mechanisms for collapse of the building under fire. However, extra bending moment due to fire will further decrease the buckling load of the columns making them more vulnerable. It is also noticed that edge and corner columns that are not in direct contact with the fire exhibit bending moment around 70% of those in contact with fire, while inner columns exhibit around 30% of the bending moment of those in contact with fire. This indicates that the fire will significantly increase the bending moment of the columns during the fire.

Figure 7.12 Model of Cardington fire test in ANSYS. (Screenshot reprinted with permission from ANSYS, Inc.)

194

Fire Safety Design for Tall Buildings

Figure 7.13 Bending moment of columns under fire loading.

Therefore, it can be concluded that fire will further decrease the buckling load of the column, and the corner and edge columns are more vulnerable during the fire. Fire protection should be made to columns especially at the edge and corner of the building to enhance the stability of the whole building. 7.4.4.3 Effect of the slenderness ratios The investigation by Burgess et al. (1992) found that columns with different slenderness ratios behave differently in fire. Stocky and slender columns perform better than those with intermediate slenderness ratios. Although residual stresses have been shown to influence the failure, their effect is no greater than at ambient temperature. Local buckling becomes more significant as a mode of failure in fire.

7.4.5 Behavior of connections As explained in Chapter 4, in the Cardington test, the excessive plastic strain in the beams due to local buckling at connection zone produces huge tensile at the cooling stage, resulting in connection failures in both partial depth endplate and fin plate connections.

Preventing fire-induced collapse

195

7.4.6 Behavior and failure mechanism of concrete column in fire Research by Dimia et al. (2011) shows that a failure during the cooling phase of a fire is more dangerous, that a failure of the structure is still possible when the fire has been completely extinguished. It has been shown that the most critical situations with respect to delayed failure arise for short-hot fires and for columns with low slenderness (short length and/or massive section). Rapid cooling of the gas temperature increases the probability for the column to survive the fire. 7.5 WHOLE-BUILDING BEHAVIOR OF TALL BUILDINGS IN FIRE

7.5.1 Research of Fu (2016b) Fu (2016b) simulated the whole-building behavior of a tall building using 3D finite element method. Based on the modeling observations and analysis of data, the following findings were made: • ‘Strong column and weak beam’ is a design principle widely used to prevent early collapse of the buildings, which enables the beams to fail earlier than columns. However, under fire, due to the thermal expansion and the strength degradation of the material, the columns are prone to fail earlier than the beams, so this design principle is not applicable. Therefore, the effective way to prevent the building collapse is to prevent the early failure of the columns. • Due to thermal expansion, the corner slabs are more vulnerable to fail than the slab at other locations. • Due to the global deformation of the whole floor plate and the restrain from the supporting steel beams, tensile membrane and compressive membrane are developed in the slab. This membrane action will enhance the load-carrying capacity of the slab. However, it is also noticed that with the increasing of temperature, the axial force of the slab dropped, which is due to the large deformation of the supporting beams.

7.5.2 Twin Towers (WTC1 and WTC2) The investigation of NIST (2005) found that the aircraft hit the towers at a high speed impacting between the 93rd and 99th floors, and did considerable damage to principal structural components (core columns, floors, and perimeter columns) that were directly impacted by the aircraft or

196

Fire Safety Design for Tall Buildings

associated debris. However, the towers withstood the impacts and would have remained standing were it not for the dislodged insulation (fireproofing) and the subsequent multi-floor fires. 7.5.2.1 Structural framing for WTC1 As shown in Figure 7.14, rather than using the traditional metal decking composite slab in the floor system, WTC1 adopted a so-called composite truss floor system due to its extra-long span of the floor. As explained in Chapter 2, the bowing of the composite floor becomes one of the key reasons for the collapse of WTC1. Two-dimensional FEM models were built by Flint et al. (2007) and Usmani et al. (2003) to investigate the collapse of the World Trade Centre towers. They found that tensile membrane action occurred by floor deflection caused by thermal expansion, with inward pulling of the exterior columns, which led to the formation of plastic hinges in those columns at floor levels. 7.5.2.2 Reason for the collapse of WTC1 As shown in Figure 2.10, Chapter 2, WTC1 and WTC2 used so-called perimeter frame-tube system with excellent robustness, and the large size of the buildings helped the towers withstand the impact. The structural system redistributed loads from places of aircraft impact, avoiding larger-scale damage upon impact. The hat truss was intended to support a television antenna and prevented earlier collapse of the building core. In each tower,

Figure 7.14 Composite truss floor system of Word Trade Center. (Reprinted from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center disaster, final report on the collapse of the World Trade Center Towers. (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1, Figures 1–6, p. 10. https:// doi.org/10.6028/NIST.NCSTAR.1.)

Preventing fire-induced collapse

197

a different combination of impact damage and heat-weakened structural components contributed to the abrupt structural collapse. As explained in Chapter 2, in WTC1, the fire weakened the core columns and caused the floors on the south side of the building to sag. The floors pulled the heated south perimeter columns inward, reducing their capacity to support the building above. Their neighboring columns quickly became overloaded as columns on the south wall buckled. The top section of the building tilted to the south and began its descent.

7.5.3 WTC7 In WTC7, according to NIST NSCTAR1A (2008), the fire lasted only on floors 7–9 and 11–17 until the building collapsed. 7.5.3.1 Structural framing for WTC7 As shown in Figure 7.15, the plan layout shows the framing of WTC7; it is primarily composed of five major frames (columns 1–42, 58–79, 59–80, 60–81, and 15–28 in the horizontal direction). As shown in Figure 7.16, there are several two-storey high trusses between floors 5 and 7. The cantilever transfer girders are also used in the vertical direction making the whole structural framing quite unique.

56

1

53

50

42

44

47

40

5

79 58

61

64

67

70

73

62

65

68

71

74

77

63

66

69

72

75

78

76 35

59

80

10 60

81 30

14A

15

18

21

24

26

28

Figure 7.15 The floor layout of WTC7. (Adapted from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center disaster, final report on the collapse of the World Trade Center Towers. (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1A, Figures 1–5, p. 6. https://doi.org/10.6028/NIST.NCSTAR.1a.)

198

Fire Safety Design for Tall Buildings

Figure 7.16 Key structural system of WTC7. (Reprint from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center disaster, final report on the collapse of the World Trade Center Towers. (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1A, Figures 1–6, p. 6. https://doi.org/10.6028/NIST. NCSTAR.1a.)

NIST (2008) made simulation of the collapse of the building using LS-DYDA. As shown in Figure 7.17, the collapse is triggered by the buckling of column 76 between levels 7 and 9. The splice of column 76 failed in bending, led to floor failure and bending failure of adjacent columns, causing a progressive failure. It is worth noting that the trusses in levels 5 and 7, which are in certain degree similar to the composite truss floor system in WTC1, have a larger stiffness; therefore, when a column fails, it pulls the adjacent column, causing continuous failure of all columns. 7.5.3.2 Reason for the collapse of WTC7 To clearly investigate the collapse mechanism of WTC7 in fire, it was simulated by the author using Abaqus® as shown in Figure 7.18. It can be seen that the exterior columns buckled at the lower floors (between floors 7 and 14) due to load redistribution to the exterior columns from the building core. The interior columns buckled due to fire. The entire building above

Preventing fire-induced collapse

199

Figure 7.17 Vertical progression of failure simulated by LS_DYNA. (Reprint from National Institute of Standards and Technology (2005), Federal building and fire safety investigation of the World Trade Center disaster, final report on the collapse of the World Trade Center Towers. (U.S. Department of Commerce, Washington, DC), NIST NCSTAR 1A, Figures 3–10, p. 40. https://doi.org/10.6028/NIST.NCSTAR.1a.)

the buckled-column region then moved downward in a single unit, and as observed, the global collapse was triggered. The investigation of NIST NCSTAR1A (2008) also shows that • Long-span floor system experiences significant thermal expansion and sagging effects. • Connection designs (especially shear connections) cannot accommodate thermal effects. • Floor framing induces asymmetric thermally induced (i.e., net lateral) forces on girders. • Shear studs fail due to differential thermal expansion in composite floor systems.

7.5.4 Cardington test In the Cardington test, the structural integrity of the composite frame was maintained during both the heating and cooling phases. This is due to several reasons.

200

Fire Safety Design for Tall Buildings

Figure 7.18 Abaqus® model showing the collapse of WTC7 building. (Abaqus® screenshot reprinted with permission from Dassault Systèmes.)

7.5.4.1 Severity of the fire As introduced in Chapter 2, in Cardington, the fire was only set in certain area, rather than spread into the whole floor plate or floor above or underneath, making less server fire than WTC1 and WTC7. 7.5.4.2 Structural framing As shown in Figure 2.16, the structural framing of Cardington is more regular compared to WTC1 and WTC7 which are kind of special. The structural framing of both WTC1 and WTC7 has been introduced in the preceding sections, and it can be seen that compared to WTC1 and WTC7, the spacing of the columns in Cardington is much smaller, and therefore less span of the slabs, making the membrane action, can be easily developed which helps to resist the collapse of the building. However, due to the unique flooring system of WTC1 and WTC7, membrane action did not stop the collapse of the building. And as it has been explained, the

Preventing fire-induced collapse

201

long-span slabs in both WTC1 and WTC7 partially contributed to the collapse of the two buildings.

7.5.5 Other research in whole building behavior A 2D finite element model for a steel moment frame with 38 storeys was analyzed by Garlock and Quiel (2007) using real fire scenarios, including a temperature–time curve consisting of heating and cooling stages. Exposure to fire was restricted to a single exterior frame bay, with vertical diffusion of the fire from the 22nd to 30th floors. According to the observations, lateral deformation of the perimeter columns occurred as a result of the fact that the heated beams underwent expansion during the stage of heating. The risk of structural collapse was increased with the formation of plastic hinges that occurred upon the beams and columns reaching the limit state under the joint action of axial forces and bending moments. Another important observation was that the perimeter column stability was dependent on the beams that braced those columns, they contributed more significantly to protection against fire. Lange et al. (2012) used two-dimensional finite element models to analyze a 12-storey composite steel frame with interior core and perimeter moment frame in fire. It is found that as the heated floors became thermally expanded, the perimeter columns were pushed outward at first, while the major deflection and catenary action of the floors subsequently caused the columns to be pulled inward. In the end, the building collapsed gradually due to the buckling of the heated perimeter columns. The progressive collapse of a typical super-tall RC frame-core tube building exposed to extreme fires was simulated by Lu et al. (2017). It is found that 1. The progressive collapse may originate in the fire-affected areas. However, due to the large redundancy of the super-tall RC frame-core tube building, the progressive collapse of the residual structure outside of the fire-affected areas can be prevented via alternative load paths. 2. The progressive collapse of the super-tall building was triggered by the flexural failure of the peripheral columns, due to the outward push by the thermal expansion of the upper floors and the inward contraction of the lower floors. 3. The more stories subjected to fire, the greater the possibility of progressive collapse. Three-dimensional FEM models were employed by Agarwal and Varma (2014) to investigate a steel moment frame building. Similar to the research results of Fu (2016b), they also found that columns had the highest probability of undergoing failure first, if the fire safety level was identical for every structural element. When the gravity columns fail, load redistribution to neighboring columns is induced by catenary and flexural action.

202

Fire Safety Design for Tall Buildings

7.6 OVERALL BUILDING STABILITY SYSTEM DESIGN FOR FIRE

7.6.1 Bracing system In tall buildings, all parts of the bracing system should be fire protected to ensure appropriate fire resistance. To ensure the overall stability of the buildings, another strategy is to place the bracing inside a protected shaft such as stairwell. However, it should ensure that the lift shaft has sufficient fire resistance.

7.6.2 Core wall design In the history of tall building design, steel core was dominant in most of the steel-framed tall buildings, such as WTC1 and WTC2. The lessons from WTC collapse show that sole steel material is not advised. Therefore, since the 9/11, the steel core is seldom used in tall buildings. All the tall buildings have started using concrete core as it is introduced.

7.7 METHODS FOR MITIGATING COLLAPSE OF BUILDINGS IN FIRE From the above investigation, it can be seen that the structural framing plays an important role in building collapse when fire starts. Special attentions should be paid to the columns during the design, especially to the edge and corner columns. Therefore, the following measures can be recommended to prevent the collapse of tall buildings in fire: 1. The collapse of WTC1 and WTC7 and the research of Fu (2016b) and other researchers all show the important role of columns in the collapse of the buildings. And the collapse of all these three buildings was triggered by the edge columns. The research of Fu (2016b) and Agarwal and Varma (2014) shows the high likelihood of early failure of columns in building fire. Therefore, prevent early failure of the columns of tall building in fire is an effective way to prevent fire induced collapse. Apart from fire protections, this can be achieved through increasing the out-plan moment capacity and shear capacity of the column, such as increasing the flange and web thickness. 2. In designing the protection regimes, special attention should be made to corner and edge columns. They are experiencing large out-plane bending due to the thermal effect, as it is explained in the simulation results of both Model 2 and Cardington tests. It is suggested to increase the thickness of the fire protection.

Preventing fire-induced collapse

203

3. Most of the connections failed due to the thermal expansion and subsequent contraction of the steel beams at the cooling stage. A robust connection with flexible deformation is needed for the sake of preventing collapse of the buildings. 4. Protected edge beams have a significant effect on the fire resistance of the structure. 5. Structural systems are also key to preventing progressive collapse. A good structural framing can provide better resistance to the effects of thermal expansion on the structural system. The lessons from WTC1 and WTC7 show that avoiding long spanning floor system and increasing the redundancy of the structure can reduce the chance of collapse. In addition, using a concrete core rather than a steel core as a major lateral stability system would greatly reduce the chance of collapse. 6. Good fire safety design in turn will reduce the likelihood of building collapse. It includes the following: • Good thermal insulation to limit heating of structural steel and to minimize both thermal expansion and weakening effects is important to prevent the collapse of a building currently. • Sufficient compartmentation to limit the spread of fires. • Automatic fire sprinkler systems with independent and reliable sources for the primary and secondary water supplies. • Thermally resistant window assemblies which limit breakage, reduce air supply, and retard fire growth. 7. As discussed in Chapter 2, NIST NCSTAR (2008) recommends a performance-based design for fire resistance that can effectively prevent collapse of a building in fire. In prescriptive design methods, the fire rating did not prevent the collapse of building. Therefore, using performance-based design is essential in preventing the collapse of the tall buildings in fire.

REFERENCES Agarwal, A., Varma, A. H. (2014), Fire induced progressive collapse of steel building structures: The role of interior gravity columns. Engineering Structures, 58, pp. 129–140. Bailey, C. G., Moore, D. B. (2000a), The structural behaviour of steel frames with composite floor slabs subject to fire: Part 1: Theory. The Structural Engineer, 78(11), pp. 19–27. Bailey, C. G., Moore, D. B. (2000b), The structural behaviour of steel frames with composite floor slabs subject to fire: Part 2: Design. The Structural Engineer, 78(11), pp. 28–33. Burgess, I. W., Olawale, A. O., Plank, R. J. (1992), Failure of steel columns in fire. Fire Safety Journal, 18, pp. 183–201.

204

Fire Safety Design for Tall Buildings

Dimia, M., Guenfoud, M., Gernay, T., Franssen, J. (2011), Collapse of concrete columns during and after the cooling phase of a fire. Journal of Fire Protection Engineering, 21(4), pp. 245–263. Flint, G., Usmani, A., Lamont, S. (2007), Structural response of tall buildings to multiple floor fires. Journal of Structural Engineering, 133(12), pp. 1719–1732. Fu, F. (2016a), Structural Analysis and Design to Prevent Disproportionate Collapse. CRC Press. ISBN 978-1-4987-8820-5. Fu, F. (2016b), 3D finite element analysis of the whole-building behavior of tall building in fire. Advances in Computational Design, 1(4), pp. 329–344. Garlock, M. E. M., Quiel, S. E. (2007), The behavior of steel perimeter columns in a high-rise building under fire. Engineering Journal, AISC, 44(4), pp. 359–372. Jiang, J., Li, G.-Q., Usmani, A. (2014), Progressive collapse mechanisms of steel frames exposed to fire. Advances in Structural Engineering, 17(3), pp. 381–398. Lamont, S., Usmani, A. S., Gilliec, M., (2004), Behaviour of a small composite steel frame structure in a ‘‘long-cool’’ and a ‘‘short-hot’’ fire. Safety Journal, 39, pp. 327–357. Lange, D., Röben, C., Usmani, A. (2012), Tall building collapse mechanisms initiated by fire: Mechanisms and design methodology. Engineering Structures, 36, pp. 90–103. Lu Xinzheng, Li Yi, Guan Hong, Ying Mingjian (2017), Progressive collapse analysis of a typical super-tall reinforced concrete frame-core tube building exposed to extreme fires, Fire Technology, 53, pp. 107–133. O’Connor, M. A., Kirby, B. R., Martin, D. M. (2003, January), Behaviour of a multistorey composite steel framed building in fire. The Structural Engineering, pp. 27–36. Nguyen, T., Tan, K. (2017), Behaviour of composite floors with different sizes of edge beams in fire. Journal of Constructional Steel Research, 129, pp. 28–41. NIST NCSTAR 1A (2008), Final report on the collapse of world trade center building 7, National Institute of Standards and Technology, US Department of Commerce. NIST NCSTAR (2005, December), Federal building and fire safety investigation of the World Trade Center disaster, final report of the National Construction Safety Team on the collapses of the World Trade Center Towers. NZS 3404 Parts 1 and 2 (1997), Steel Structures Standard. Usmani, A. S., Chung Y. C., Torero, J. L. (2003), How did the WTC towers collapse: A new theory. Fire Safety Journal, 38(6), pp. 501–533. Wong, B., Burgess, I. (2013), The influence of tensile membrane action on fireexposed composite concrete floor-steel beams with web-openings. Procedia Engineering, 62, pp. 710–716. Lin, S., Huang, Z., and Fan, M. (2015). The effects of protected beams and their connections on the fire resistance of composite buildings. Fire Safety Journal, 78, pp.31–43.

Chapter 8

New technologies and machine learning in fire safety design

8.1 INTRODUCTION There are many new technologies developed recently which can be used for fire safety design, such as PAVA system, IOT, and smart building management system (BMS). They will be introduced in this chapter. Machine learning (ML) is a new technology that is gradually replacing human beings in most of the disciplines. Their applications in construction industry are still restricted. However, it will have a great impact in all areas of the global economy in the future. It will change the construction industry fundamentally. Therefore, it is beneficial to the engineers to have some basic ML knowledge. For this purpose, some pilot studies of using machine leaning in fire safety design will also be introduced in this chapter.

8.2 NEW TECHNOLOGIES IN FIRE SAFETY With the fast development of the technology, some new technologies such as PAVA, IOT, and smart BMS have been developed. They have the great potential to enhance current fire safety design measures and bring the fire safety design to a new level.

8.2.1 PAVA alarm systems Personal Announcement Voice Activated (PAVA) alarm systems integrate alarm detection systems to controlled evacuation of buildings by means of clear pre-recorded spoken messages rather than bells or sounders. In this way, the system automatically broadcasts specific messages to alert occupants in case of a danger and direct them to the nearest safe exits or refuge areas. Traditional bells and sounders only give a warning, but they do not indicate the nature of the emergency. This may leave people uncertain, and often such alarm signals are ignored—potentially with fatal consequences. A voice alarm system provides clear easily understood instructions 205

206

Fire Safety Design for Tall Buildings

via pre-recorded messages, ensuring that even untrained personnel can be evacuated speedily and efficiently. Therefore, as it reduces the reaction time very effectively, it is especially effective for complex buildings. Phased evacuation can be assisted by PVAV using a combination of clear pre-recorded messages and live announcements to enable occupants in selected areas to be evacuated in turn. The voice alarm system works automatically, with all controls easily overridden by fire officers or building control when needed.

8.2.2 IOT in fire safety Internet of Things (IOT) has been gradually used in fire safety design recently. The IOT usually refers to the idea that everyday objects could be connected to the internet. 8.2.2.1 Fire safety sensors and BMS IOT sensors are now used to refer to any small, internet-connected devices which record specific data and transmit this to a central location or device to be interpreted. These sensors might record audio, video, temperature data, location data, and much more besides. Smart buildings are properties which are controlled in part by autonomous computer software, known as BMS. These sensors can maintain a specific temperature in different rooms, turn lights on and off, and perform other tasks which benefit from external data. Special heatproof sensors could detect the temperature of fires, giving firefighters a clue of intensity of fires, allowing them to alter their equipment and approach. IOT sensors could show not only the location of a fire started but also the speed and direction of its spread. All of this information could be transmitted automatically to fire crews, even happening alongside the emergency call. The emergency calls will be automated by the BMS, with the system forwarding vital data to the local fire department’s computer systems, which could then organize their own proportionate response. IOT can work together with voice alarm system to inform occupants the best escape routes, based on the direction of the fire spread. 8.2.2.2 Fire suppression IOT technology can link a fire alarm or carbon monoxide detector with home appliances. If the system detects the presence of fire or carbon monoxide, it can automatically shut off these ignition sources. IOT can also work together with sprinkler systems with more targeted firefighting capabilities,

New technologies and machine learning

207

helping to put out small fires and stem the tide until emergency crews arrive. By sensing exactly where the fire is, the nature of the fire, and whether there are any occupants in the room, a smart IOT-enabled fire system can choose different measures to specific rooms.

8.3 MACHINE LEARNING IN FIRE SAFETY DESIGN ML is a part of AI technology which evolved from the study of pattern recognition and computational learning theory in artificial intelligence (AI). ML algorithms build a mathematical model based on sample data, known as “training data,” which make predictions and decisions. It is closely related to computational statistics. It uses statistical techniques to let computers learn. ML uses complex models and algorithms to predict. Therefore, the core of ML is to explore the study and construction of algorithms that can learn and make predictions on existing data. The study of mathematical optimization delivers methods, theories, and application domains to the field of ML. Data mining is another related field of study to ML. These analytical models allow machine to decisions through learning from historical relationships and trends in the data. Therefore, ML has been used widely in almost all the industries recently, such as finance, security, and medical service. In the past several years, ML has been widely used in many industries. Nowadays, in certain countries, ML has overtaken even experienced doctors in preliminary diagnosis of diseases. It is also widely used in finance (such as stock price prediction), cyber security (ML is used to predict the possible cyber-attacks), etc. Compared with a human being such as a data analyst or a doctor, the great advantage of AI lies in its capacity to process and explore extraordinarily large dataset and therefore, makes a much more accurate prediction than experienced professionals. The combination of the AI and Big Data supported by cloud computing further improves the capacity of AI. This is because AI can analyze a huge amount of Big Data only with the help of the large computational power provided by cloud computing. Therefore, the application of AI in the construction industry is one of the promising approaches for providing efficient and cost-effective design solutions. Using the machine to analyze large scope of scenarios and predict the accurate failure pattern of building under fire would be one of the promising solutions for performing an effective fire safety design of the building. The development of a new technique based on ML would enable a quicker assessment of a structure’s vulnerability to fire and bring it within consideration for all steel-framed projects. Although construction research has considered ML for more than two decades, it had rarely been applied to fire safety design of buildings (Fu, 2018, 2020).

208

Fire Safety Design for Tall Buildings

8.3.1 Machine learning and its application in the construction industry However, as a traditional industry, AI and its applications in construction industry are far behind other areas. Although construction research has considered ML for more than two decades, it has rarely been applied to structural safety design. Some research has been undertaken in the past by using the ML for certain construction problems. Puri et al. (2018) used ML to predict the SPT N-value of soil using. Paudel et al. (2015) used ML for the prediction of building energy demand. Zhang et al. (2018) developed an ML framework for assessing post-earthquake structural safety. Shi et al. (2018) set up an evaluation model to assess the intelligent development of 151 cities in China using back-propagation neural network theory. Tixier et al. (2018) used Random Forest (RF) and Stochastic Gradient Tree Boosting (SGTB) methods to predict the injury in the construction sites. Özturan et al. (2018) used the artificial neural network to predict the concrete strength. The primary problem that exists in the above application is that most of them can only solve simple problems such as prediction of the injury and relationships of the in-place density of soil or prediction of the strength of the concrete. In addition, few applications to fire safety design have been found. Therefore, further research on ML in solving more complex construction problems is imperative.

8.3.2 Problems experienced in the conventional structural fire analysis approach As introduced in the preceding chapters, one of the primary fire safety design measures is to place fire protection on the structural members. As introduced in Chapter 5, in the project of the Shard, all primarily beams are protected for at least 60 min. However, due to limitations in understanding the accurate response and failure patterns of the buildings, there is a possibility of overdesign due to addition of fire protection to unnecessary structural members. In certain cases, unnecessarily thick fire protection is used, which results in huge waste. Although an excess of money is spent due to overdesign of the structural fire protection, in certain circumstances, the building may not sufficiently be safe because of the wrong fire protection regime chosen. The above reported issues are primarily due to the limited capacity of an engineer to fully understand the failure pattern of a building in the fire conditions. On the other hand, the present deterministic methods for assessing the fire protection regimes are time consuming. This is because fire development and subsequent structural response depend upon numerous factors. The appraisal of structural response in fire is challenging given the sources of uncertainty. In addition, for a structure such as a tall building, the

New technologies and machine learning

209

structure systems are much more complicated, which also brings additional difficulties to the structural fire analysis. The traditional design process is therefore time consuming and is limited by the ability of an engineer to fully understand the failure potential of the structure under different fire loadings. One of the best solutions to tackle above issues is to use ML.

8.3.3 Predicting failure patterns of simple steel-framed buildings in fire Fu (2018, 2020) developed a new ML framework for fast prediction of the failure patterns of simple steel-framed buildings in fire and a potential assessment of subsequent progressive collapse. Critical temperature method introduced in Chapter 4 is used to define the failure patterns of each structural member. The training set of failure patterns is generated using both the Monte Carlo simulation and random sampling, which can guarantee a robust and sufficient large dataset for training and testing, thus guarantying the accurate prediction. Three classifiers are chosen for prediction of failure patterns of buildings under fire: Decision Tree, k-nearest neighbor (kNN), and Neural Network using Google Keras with TensorFlow which is specially used for Google Brain Team. The ML framework is implemented using codes programmed by the author in VBA and Python language. A case study of a two-storey by two-bay steel-framed building was made. Two different fire scenarios were chosen. The procedure, shown in Figure 8.1,

Figure 8.1 ML framework for the fire safety assessment of buildings (Fu, 2018).

210

Fire Safety Design for Tall Buildings

gives satisfactory prediction of the failure pattern and collapse potential of the building under fire. 8.3.3.1 Define failure pattern To enable ML, the failure patterns of the structure need to be first classified. There are different failure patterns when buildings are in fire. They have been introduced in the preceding chapters. Due to the complexity, it is difficult to digitalize and quantify these failure patterns to make the machine understand. One possible solution is to digitalize the failure patterns and subsequent image recognition, but it will be computationally difficult. However, in a fire safety design, the primary job for an engineer is to judge whether a member will fail. Therefore, a simplified critical temperature method introduced in Chapter 4 is used by Fu (2010, 2018) to define the failure patterns of each structural member. 8.3.3.2 Dataset generation using the Monte Carlo simulation and random sampling To accurately predict the failure mode of one type of structure, the computer should be trained with a vast database of sufficient failure cases due to fire from real projects. The amount of the training cases is crucial for the accurate prediction result of the machine. However, there is no sufficient records of fire incidents available across the world; therefore, it will be hard to find sufficient training cases in the construction industry for the fireinduced failure modes. To tackle this problem, Fu (2018, 2020) developed a novel method based on the Monte Carlo simulation, and random sampling is developed to generate sufficient large dataset in this project. The key variables that affect the failure patterns of the structural members under fire are generated using the Monte Carlo simulation, such as opening factors and fire load density (for fire load), imposed load (for gravity load), and steel grades for different structural members. After the Monte Carlo simulation, these parameters are selected using random sampling techniques with equal opportunities for structural fire analysis based on the Eurocode, and failure judgment can also be made using the so-called critical temperature method available from Eurocode. 8.3.3.3 Training and testing After the dataset is generated, the computer shall be trained and tested using the dataset. Data scientists normally use 80% of the data for training and 20% of the data for testing.

New technologies and machine learning

211

8.3.3.4 Failure pattern prediction If the testing results are satisfactory, the dataset generated is used for prediction. Different algorithms can be used for learning and prediction. Fu (2018, 2020) used kNN, Decision Trees, and Neural Network. Both Neural Network and kNN provide satisfactory prediction results. 8.3.3.5 Fire safety design and progressive collapse potential check based on prediction results Based on the prediction of each single member, the collapse potential of the building can be checked using the member removal method specified by GSA (2003) and DoD (2009) the collapse potential can be assessed. Based on the prediction of the response of each structural member by the machine, a viable fire protection scheme can be selected.

8.3.4 Predicting and preventing fires with machine learning Professor Jae Seung Lee and his students at Hongik University studied the information held by the city’s fire service. Using ML, the university team was able to predict the probability of fires with an impressive 90% accuracy. The data supplied to the algorithm told students which neighborhoods were most at risk, allowing the Seoul Metropolitan Fire and Disaster Management Headquarters to deploy firefighters and patrol in the most vulnerable regions.

8.3.5 Machine learning of fire hazard model simulations for use in probabilistic safety assessments at nuclear power plants Worrell et al. (2019) explored the application of ML to generate accurate and efficient metamodels for probabilistic fire safety assessments of nuclear power station. The process involved fire scenario definition, generating training data by iteratively running the fire hazard model called CFAST over a range of input spaces using the RAVEN software. The input and output data from a population of 675,000 CFAST were consolidated into a single comma-separated variable (.csv) file. R software was used for final metamodel selection and tuning. They used both Decision Tree and kNN as prediction algorithms. It is found that a kNN model fit the vast majority of calculations within ±10% for maximum upper-layer temperature and its timing. The resulting kNN model was compared to an algebraic model typically used in fire probabilistic safety assessments. This comparison illustrated the potential of

212

Fire Safety Design for Tall Buildings

metamodels to improve modeling realism over simpler models selected for computational feasibility. While the kNN metamodel is a simplification of the higher-fidelity model, the error introduced is quantifiable and can be explicitly considered.

8.3.6 Learning algorithms and programming language 8.3.6.1 Learning algorithms Up to date, three learning algorithms—Decision Tree, kNN, and Neural Network—are widely used for fire safety-related engineering problems. The research by Fu (2020) shows that kNN and Neural Network are the two promising classifiers for this particular engineering problem. Decision Tree yields less promising results. Worrell et al. (2019) also show the accuracy of kNN in their research. 8.3.6.2 Programming language Python is one of the most popular languages for ML. It is available in various compliers such as Anaconda. It has most of the popular learning algorithms pre-programmed. Therefore, they are very handy for coding. Python libraries contain pre-written codes that can be imported into a code using Python’s import feature. A Python framework is a collection of libraries intended to build a model of ML easily, without having to know the details of the underlying algorithms. An ML developer, should know how the algorithms work in order to know what results to expect, as well as how to validate them. MATLAB is another widely used ML language package. Engineers and other domain experts have deployed thousands of ML applications using MATLAB. R is an open-source language, so people can contribute from anywhere in the world. In R, the Black Box is referred to as a package. The package is nothing but a pre-written code that can be used repeatedly by anyone. There are many ML packages in R for a programmer to use.

REFERENCES Fu, F. (2018), Fire safety assessment of buildings through machine learning, MSc thesis, University of Oxford. Fu, F. (2020), Fire induced progressive collapse potential assessment of steel framed buildings using machine learning. Journal of Constructional Steel Research, 166, pp. 105918–105918.

New technologies and machine learning

213

GSA (2003), Progressive collapse analysis and design guidelines for new federal office buildings and major modernization projects. The U.S. General Services Administration. ISO (1984), Paints and Varnishes—Colorimetry. Part 1: Principles. ISO 7724-1. ISO, Geneva. Özturan, M., Kutlu, B., Özturan, T. (2008), Comparison of concrete strength prediction techniques with artificial neural network approach. Building Research Journal, 56, pp. 23–56. Paudel, S., Nguyen, P. H., Kling Wil, L., Elmitri, M., Lacarri`ere, B., et al. (2015). Support Vector Machine in Prediction of Building Energy Demand Using Pseudo Dynamic Approach. Proceedings of ECOS 2015 – The 28th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, Jun 2015, Pau, France. hal-01178147. Puri, N., Prasad, H. D., Jain, A. (2018), Prediction of geotechnical parameters using machine learning techniques. Procedia Computer Science, 125, pp. 509–517. Shi, H. B., Tsai, S. B., Lin, X. W., Zhang, T. Y. (2018), How to evaluate smart cities’ construction? A comparison of Chinese smart city evaluation methods based on PSF. Sustainability, 10(1), 37. Unified Facilities Criteria, Department of Defense, UFC 4-023-03 (2003), Design of Buildings to Resist Progressive Collapse 14 July 2009 with Change 2n. Worrell, C., Luangkesorn, L., Haight, J., Congedo, T. (2019, March), Machine learning of fire hazard model simulations for use in probabilistic safety assessments at nuclear power plants. Reliability Engineering & System Safety, 183, pp. 128–142. Zhang, Y., Burton, H. V., Sun, H., Shokrabadi, M. (2018), A machine learning framework for assessing post-earthquake structural safety. Structural Safety, 72, pp. 1–16.

Chapter 9

Post-fire damage assessment

9.1 INTRODUCTION Post-fire damage assessment is one of the key methods for retrofitting and reconstructing the structure after fire. Different damage assessment techniques including destructive and nondestructive assessment methods for concrete and steel structures will be introduced in this chapter. 9.2 POST-FIRE DAMAGE ASSESSMENT Assessment of fire-damaged structures is essential because while the damages are often open to see, the structural damages that affect the elements and structures that provide support can’t easily be seen. Consequences of fire damage can include dramatically reduced strength of the steel reinforcement, prestressing or post-tensioning tendons; partial or total loss of the strength of concrete; and delamination of the concrete cover, which reduces its ability to protect the embedded steel. There are several methods available for both concrete and steel structures, which will be introduced in this section.

9.2.1 Post-fire damage assessment of concrete structure 9.2.1.1 Visual inspection The bearing capacity of the vertical structural elements (such as columns and shear walls) is critically important for the stability of a structure, and so they, as well as beams and slabs, should be visually inspected at the first instance. 9.2.1.1.1 Damage to columns If the concrete cover had fallen from columns, the longitudinal reinforcement or stirrups should be checked for any damage. 215

216

Fire Safety Design for Tall Buildings

9.2.1.1.2 Damage to shear walls Spalling can be usually overserved in concrete walls, so it should be checked if there is any disengagement between the concrete and the reinforcement. 9.2.1.1.3 Damage to beams Any decrease in the bearing capacity of beams will cause excessive deflections on the elements. In the event of fire, the bond between concrete and longitudinal reinforcement deteriorated, and hence the tension strength of longitudinal reinforcement decreases, which leads to large deflection of the beams. Therefore, the excessive deflection of the beams can be an indication of beam failure. 9.2.1.2 Schmidt rebound hammer A Schmidt hammer can be used to determine estimated equivalent cube strengths of concrete members and assess the differences between fire-damaged and unaffected areas. Schmidt hammer, shown in Figure 9.1, is a device used to measure the elastic properties and strength of concrete, mainly surface hardness and penetration resistance. The hammer measures the rebound of a springloaded mass impacting against the surface of a sample. The test hammer hits the concrete at a defined energy. Its rebound is dependent on the hardness of the concrete and is measured by the test equipment. By referring to a conversion chart, the rebound value can be used to determine the concrete’s compressive strength. 9.2.1.3 Petrographic analysis Petrographic analysis can determine the effects and extent of fire damage on a microscopic level. In petrographic analysis techniques, cast thin section, X-ray diffraction, and scanning electron microscope are the most widely used; electronic probe analysis is also sometimes used. 9.2.1.4 Spectrophotometer investigations After fire, samples of concrete can be taken from the building for spectrophotometer testing. Color alternation in concrete is noticed, which can be used to judge temperature levels and the damage caused by the fire. The hue, value, and chroma of each sample can be measured by spectrophotometer according to ISO (1984). The results of this test can be used to find the depth of the concrete affected by the fire and, in turn, the thickness that should be repaired.

Post-fire damage assessment

217

Figure 9.1 Schmidt rebound hammer. (This file is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license. https://upload.wikimedia.org/ wikipedia/commons/1/1b/Schmidt_hammer_testing.jpg.)

9.2.1.5 Reinforcement sampling Reinforcement samples are taken from the buildings after fire for yield strength and metallurgical testing. Thus, the residual strength of the reinforcement is determined. 9.2.1.6 Compression test The strength values of the damaged concrete are obtained through direct compression test on the samples taken from the building exposed to fire. It is known that the compressive strength of the concrete not only degrades with temperature rise but also changes due to rate of heating, duration of the fire, loading, type of the aggregate, and water–cement ratio.

218

Fire Safety Design for Tall Buildings

9.2.2 Post-fire damage assessment of structural steel members 9.2.2.1 Methods for post-fire damage assessment The method for testing the residual strength of steel after fire includes on-site coupon tensile testing, chemical composition analysis method, and surface hardness method. The most accurate one is on-site coupon tensile testing. In this method, the coupon is cut from the structural members of the building to perform tensile testing. However, this method causes different degrees of damage to the structure, which may make post-fire restoration work difficult. The chemical composition analysis method also needs on-site sampling, and the process is tedious. 9.2.2.2 Nondestructive post-fire damage assessment of structural steel members using the Leeb harness method For steel structure, surface hardness methods (including the Brinell hardness method, Rockwell hardness method, Victoria hardness method, and Leeb hardness method) are nondestructive, but only few of them have been used in post-fire damage assessment. Liu and Fu et al. (2020) developed a quick, simple, and efficient nondestructive detection method to measure the strength of steel after fire. It uses the so-called Leeb hardness method by means of establishing a relationship between the residual strength of steel members after fire and the Leeb hardness, and the postfire steel strength can be fast determined without making any damage to the structural members. As shown in Figure 9.2, the Leeb hardness testing is a nondestructive method for testing the strength of the steel members. It was invented by Dietamar Leeb in 1975. The method of the Leeb hardness testing is to drop certain weight of object through a tube to the surface of the specimens and test the impact velocity and velocity of the object at 1 mm distance from the surface when it bounces back. A digital Leeb hardness tester TIME5351 is shown in Figure 9.2. The specimens were first grinded into smooth zones (30 mm × 60 mm) for the test. The surface roughness was first assessed using roughness detectors. As shown in Figure 9.3, tests will be done for each smooth zone, and the average readings after removing the maximum and minimum values can be used. A total of 120 Chinese H-shaped steel sections were selected for testing the Leeb hardness after fire by Liu and Fu (2020). Based on the test results, a linear regression analysis was made, and the correlations between the Leeb hardness and the residual tensile strength of flange and web are shown in Figures 9.4 and 9.5, respectively.

Post-fire damage assessment

Figure 9.2 TIME5351 Leeb.

Figure 9.3 Leeb hardness testing of an I section.

219

220

Fire Safety Design for Tall Buildings

Figure 9.4 Correlation between the Leeb hardness and the residual tensile strength of flange Liu and Fu (2020).

Figure 9.5 Correlation between the Leeb hardness and the residual tensile strength of web Liu and Fu (2020).

Post-fire damage assessment

221

REFERENCES Ada, M., Yüzer, N., Ayvaz, Y., Postfire damage assessment of a RC factory building. Journal of Performance of Constructed Facilities, 33(5), pp 040190471–04019047-12. ISO (1984), Paints and varnishes—Colorimetry. Part 1: Principles. ISO 7724-1. ISO, Geneva. Liu, D., Liu, X., Fu, F., Wang, W. (2020), Nondestructive post-fire damage assessment of structural steel members using Leeb harness method. Fire Technology, 56(4), pp. 1777–1799.

Index

acceptance criteria 78, 120 active control system 72 ADINA 3, 158, 160 algorithms 207, 211, 212 ambient temperature 20, 88, 89, 92, 93, 171, 179, 188, 194 ANSYS 3, 155, 157, 158, 160, 161, 162, 163, 164, 177, 193 Approved Document B 13, 14, 15, 26, 28, 41, 79, 81, 112, 119, 122, 123, 126–130, 132, 133, 148 ASTM E119 34, 40 atmosphere temperature 22, 37, 39, 44– 46, 50, 51 150, 153, 163, 170 beam openings 109 bending moment 103, 104, 193, 194, 201 BMS 205, 206 British Research Establishment 9, 12, 14, 20– 25, 40 BS 476 31, 33, 41, 61, 69, 74, 80, 81, 106, 107, 111, 122, 147, 148, 180 BS 5839–6 131, 148 BS 8214 121, 148 BS 9999 31, 41 BS5950–8 31, 41 BSI PD7974 30 Burj Khalifa 4, 113, 137, 140, 141, 142, 147 Cardington fire test 20, 22, 23, 25, 71, 193

cavity barrier 14, 120, 123, 124 CECS 35, 41 CFAST 178, 211 CFD 4, 39, 50, 51, 149, 154, 155 China Zun Tower 2 cladding 1, 8, 9, 11, 15, 59, 133, 134 column buckling 23, 24, 92, 184, 185 compartment floors 18, 80, 119, 120, 122 compartment wall 3, 12, 18, 38, 47, 69, 78, 81, 93, 119, 120, 122–125, 129 concrete cover 88, 215 connection failure 24, 184, 194 connection 4, 18, 20, 22, 23, 24, 38, 77, 98, 100, 109, 132, 184, 186, 187, 194, 199, 203 convection 57, 61, 62, 95, 163 cooling stage 109, 150, 184, 188, 194, 201, 203 corner column 192, 193, 202 critical temperature method 4, 40, 77, 78, 88–90, 109, 110, 209, 210 Decision Tree 209, 211, 212 degradation 65, 66, 79, 149, 160, 174, 185, 193, 195 Degree of Utilisation 90 deterministic 4, 34, 113, 115, 116, 118, 144, 145, 169, 208 DoD 35, 211 EN 1991–1-2 1, 2, 5, 32, 41, 51, 52, 55, 56

223

224

Index

EN 1993–1-2 1, 2, 5, 32, 41, 63, 75, 88, 89, 90, 95, 112, 148, 155, 165, 166, 180 EN 1994–1-2 1, 2, 5, 32, 41, 63, 75, 89, 98, 112, 155, 165, 166, 180 escape routes 29, 114, 125, 126, 206 evacuation route 3, 4, 9, 13, 26, 37, 77, 113, 114, 125, 128, 131, 142, 147 EVC 131 exit route 126, 128, 142 exit 1, 14, 126, 128, 130, 131, 136, 139, 142, 147, 205 external fire spread 28 fire compartment 10, 18, 38, 45, 49, 60, 61, 79, 119, 121–124, 126, 138, 150, 151, 155, 156 fire resistance 4, 9, 14, 20, 22, 25, 27, 31–34, 39, 41, 43, 46, 47, 54, 57, 59, 69–71, 74, 75, 77, 79–81, 83, 86, 88, 93, 108, 109, 111, 112, 115, 116, 119, 121–123, 134, 138, 141, 142, 144–148, 150, 168, 169, 180, 190, 191, 202–204 failure pattern 207–211 FDS 3, 5, 131, 155, 178 finite element 3, 37, 40, 41, 94, 97, 110, 112, 146, 156–158, 164–166, 178, 180, 185, 195, 201, 204 fire alarm and communication systems 138 fire alarm system 6, 120, 138, 140 fire alarms 12 fire and smoke suppression system 134 fire development 1, 22, 43, 44, 48, 49, 157, 166, 208 fire doors 12, 14, 29, 120, 122, 129 fire fragility functions 174, 175, 180 fire lift lobby 128 fire load density 44, 46, 47, 55, 118, 150, 170, 174, 210 fire loading 53, 55, 118, 145, 167, 185, 193, 194, 209, 217 fire resistance rating 39, 70, 71, 79, 109, 116, 121, 146, 150 fire risk assessment 29, 30, 34, 117 fire safety engineering 1, 2, 30–34, 41, 42, 57, 58, 60, 64, 74, 111, 181

fire scenario 1–4, 32, 33, 41, 46, 47, 50, 53, 60, 71, 79, 113–118, 144, 145, 174, 175, 186, 189, 193, 201, 209, 211 fire severity 34, 53, 54, 70, 116, 168, 169 fire spread 20, 27, 28, 30, 32, 41, 47, 48, 53, 55–60, 64, 70, 71, 74, 75, 79, 111, 120, 126, 132, 133, 206 fire stop 18, 122–124 fire suppression 27, 136, 137, 142, 206 fire test 7, 20, 22, 23, 25, 31, 33, 34, 40, 41, 45, 54, 61, 70, 71, 73, 74, 84, 93, 107, 108, 111, 122, 147, 148, 180, 188, 193 firefight access 142 fire-induced collapse 4, 35, 77, 179, 183, 184 flashover 43, 44, 48–50, 151, 153, 155 FSO 28, 29 fuel-controlled fire 4, 22, 43, 49, 50 fully developed fire 50, 174, 178 GB50016 35, 41 GSA 35, 211, 213 heat transfer 4, 36, 37, 43, 61, 62, 94, 95, 149, 155–158, 160, 161, 165, 166, 174, 179 holistic approach 115 Housing Act 2004 28, 29 integrity of compartmentation 124, 147 internal fire spread 27 International Fire Code 33, 41, 115 intumescent paint 65, 72, 121, 124, 134 IOT 4, 205, 206, 207 ISO 16730–1 32, 41 ISO 16730–1 32, 41 ISO 24679–1 32, 42 ISO 834–1 33, 41, 71, 74, 75 isotherm 86 kNN 209, 211, 212 lateral stability 1, 18, 20, 144, 189, 203 Leeb hardness 218–220 life safety 3, 32, 113, 115, 183 limit state function 169, 171, 172

Index load ratio 88, 89, 90, 91, 100 local buckling 23, 24, 186, 194 localized fire 43, 49–52, 75, 144, 149, 150, 151, 155 long-cool fire 4, 43, 46, 49, 50, 53, 75, 189, 190, 204 LS_DYNA 3, 179, 199 machine learning 176, 180, 207, 208, 211, 212 MATLAB 172, 212 mechanical design 47, 67, 69, 88, 160 mechanical properties 36, 59, 66, 88, 160 Mega Frame 1 membrane action 101, 102–104, 188 membrane effect 22 moment capacity approach 93, 95, 97 Monte Carlo simulation 118, 145, 169, 170, 172, 174, 176, 209, 210 Neural Network 208, 209, 211–213 NIST 15–18, 39, 42, 53, 71, 75, 112, 113, 178, 179, 181, 195, 196–199, 203, 204 nondestructive 5, 215, 218 occupancy 26, 47, 48, 55, 56, 115 one-zone model 51, 150, 153, 177 opening factor 44–47, 49, 86, 115, 116, 118, 145, 170, 210 OpenSees 3, 164, 177, 178, 179, 181 overall stability 69, 169, 202 OZONE 3, 177 parametric fire temperature 44, 46, 50, 68, 150, 157, 186 passive control system 72 PAVA 4, 205 performance-based design 38, 115–117, 203 phased evacuation 130 plastic strain 69, 187, 189, 190, 194 post-fire damage assessment 215, 218 post-flashover fire 43, 150, 153 post-tension slabs 107 PP fiber 84–86 pre-flashover 43, 48, 50, 151, 155 prescriptive-based design 38, probabilistic method 4, 116, 145, 167 progressive collapse 17, 18, 201, 203, 209, 211, 212

225

protected elevators and staircases 128 protected lift 125, 128, 129, 132, 137, 147 protective layers 88 Python 209, 212 qualitative fire risk assessment 117 quantitative fire risk assessment 117 R 212 radiation 57, 58, 61, 62, 95, 163 refuge 12, 125, 126, 130, 131, 139, 141, 142, 205 reliability index 168 reliability-based method 167 scanning electron microscopic 84 Schmidt hammer 216 SCI-P288, 125 shear resistance 98 short-hot fire 4, 43, 46, 49, 50, 53, 82, 186, 190, 195 slab failure 25, 184 smoke detectors 32, 135 smoke extraction system 136 spalling 20, 82–84, 102, 216 spectrophotometer 216 sprinkler 10, 12, 18–20, 31–33, 37, 48, 55, 59, 71, 72, 117, 137, 138, 142, 172, 174, 203, 206 stairway pressurization systems 136 standard fire temperature 44, 45, 53, 60, 64, 74, 150, 157, 163 statistical parameters 118, 172, 173 steel cores 15, 26 steel temperature 22, 74, 92 Stefan–Boltzmann constant 62 structural fire design 66–68, 77 structural fire analysis 1, 3, 44, 66, 69, 114, 149, 165, 175, 208–210 structural framing 202, 203 structural response 1, 3, 30, 50, 57–60, 64, 74, 111, 115, 116, 157, 174, 179, 208 structural stability 30, 38, 77, 140, 183, 184 temperature analysis 67, 68 temperature distribution 89, 93, 97, 110, 157, 186 temperature field 37, 156, 158, 160, 163

226

Index

temperature profile 93–96, 106, 157, 185 thermal analysis 158, 161, 163, 164 thermal barrier 83, 88 thermal conductivity 49, 62, 64, 73, 95, 174, 178, 184 thermal expansion 23, 24, 53, 79, 83, 160, 186, 191, 192, 195, 196, 199, 201, 203 thermal inertia 188 thermal mechanical analysis 4, 145, 149, 158, 160, 164, 166 thermal properties 47, 150 thermal response 1, 3, 4, 31, 39, 40, 51, 59, 60, 61, 82, 88, 116, 149, 185

traveling fire 53 two-zone model 150, 151 UFC 3–600–01 35 vapor pressure 84 ventilation-controlled fire 49 whole-building behavior 195 worst-case fire 4, 50, 113, 118, 144, 145 zone model 149, 150